Hospital First to Test All Patients for Millions of Tumor Mutations
“Incredibly valuable,” is how Barrett Rollins, Chief Scientific Officer of the Dana Farber Cancer Institute, describes his hospital’s unprecedented Profile program. In that program, the largest of its kind, all adult cancer patients at two Boston hospitals—Dana Farber and Brigham Women’s—are offered some of the most extensive, next-generation-sequencing (NGS) tests in existence.
Not some. All.
“I am aware of no other genetic profiling projects designed to test every patient at a cancer center,” says Rollins, contacted by email. “The database this creates is linked to the database containing clinical information on all of these patients. It has created a well-defined cancer cohort comprised of thousands of patients.”
The program is also moving fast. It just banked its 5,000th tumor profile after only two years. And it has already upgraded its own technology.
The first Profile platform (called OncoMap) let scientists scan, for every consenting patient, tumor DNA for 471 known cancer mutations—small errors in DNA code—in 41 genes thought linked to cancer. The new platform, OncoPanel, is orders of magnitude more thorough. It uses massively parallel, NGS sequencing to decode the DNA blueprint of 305 genes, and to find potentially millions of mutations. OncoPanel, designed by Harvard and MIT, also mines for other genetic changes: deleted or amplified pieces of DNA, or broken chromosomes repaired inadequately.
Profile teams now find 20 mutations per sample, where once they found two, at most.
These highly detailed snapshots should spark revolutions on two levels. First, they give clinicians a better understanding of their patients’ unique tumors. Clinicians can now pull off the shelf, for patients, drugs they never would have tried otherwise: personalized medicine.
“In one case, a patient received four cycles of chemotherapy for a cancer type diagnosed on the basis of highly expert histopathology," says Rollins. "His tumor grew through his therapy. But when he saw us and had his new tumor growth resected, the tumor was automatically genotyped as part of Profile. A distinctive genetic rearrangement was found that is characteristic of a cancer different from the one he was being treated for. That cancer has a chance for cure using completely different chemotherapy.”
Additionally, this massive NGS effort should give the hospitals a rare birds' eye view of many shared mutations. One investigator is finding common mutations in melanoma that alter immune system responses. Another is unearthing mutations common to aggressive testicular cancers.
The team wants to make extensive sequencing routine. “The concept is that we're analyzing specimens being collected for clinical care,” Rollins says. “We do not specifically obtain specimens for testing that would not have been obtained during clinical care. The idea is to see whether we can incorporate enterprise-wide genetic testing into routine clinical practice.”
So the number of specimens available for Profile testing from a given patient depends on how many were obtained during clinical biopsy, resection, or bone marrow aspiration procedures. The team does not automatically test both patients’ primary tumors and metastases.
This matters. Research increasingly shows tumors are heterogeneous. In breast cancer, some studies find women whose primary tumors are “Her2 negative” can also have Her2 positive circulating tumor cells. Up to 30% of Phase 1 breast cancer patients may possess such cells. Other studies find bone metastases of Her2-negative patients can test Her2 positive on recurrence. Such patients would suddenly be eligible for the successful targeting drug, Herceptin.
Rollins agrees additional testing can be key. “We have studies separate from Profile that look at heterogeneity between primary tumors and mets. But for Profile, the source of the specimen (primary or met) depends on what was obtained clinically. We only have enough money to subsidize testing of one specimen per patient. We test what's been resected.”
But, he adds, “If a provider or a researcher has funds to pay for second or third Profile tests on other specimens, we will perform the tests. It's important to remember Profile is a large-scale enterprise experiment to see if we could set up a platform for universal specimen-based testing. New technologies, or new testing algorithms (like routinely testing recurrences or metastases) can be incorporated as opportunities arise.”
Nearly 18,000 patients are enrolled. All can have their physician notified if their test unveils mutations for which treatments exist. Some Boston Children’s Hospital patients are eligible.
Other hospitals are launching smaller programs. At Memorial Sloan Kettering Cancer Center (MSKCC) in New York “we are in the final stages of validating a clinical NGS test—called IMPACT—involving exon capture of all protein-coding exons of 340 cancer genes,” says MSKCC Genomics Researcher Mike Berger. “This will provide the mutation and copy number status for these genes. We are also capturing selected non-coding sequences to detect recurrent gene fusions. This would be suitable for patients with breast cancer and other solid tumors.” MSKCC has many smaller NGS projects.
A "very important" NGS program at Ottawa Hospital in Ontario Canada is “in planning stages,” says Ottawa oncologist David Stewart.
The US National Institutes of Health, via the Cancer Genome Atlas, is in the midst of a multiple institute, $100 million-a-year NGS project.