tumor type and not necessarily a specific kind of cancer, our first focus is on the clinical efficacy of the medication and where it delivers the most value to patients.”
Express Scripts, whose clients include insurance companies and large employers, is working to change cancer-drug pricing, from one price no matter what the indication, to a price set to a drug’s tissue-specific effectiveness. “If a drug extends life by six months for lung cancer patients, but only a few weeks for a colorectal cancer patient—or vice versa—we believe that payers should not have to pay the same amount,” says Henry, whose company is now the target of a proposed $67 billion buyout by insurance giant Cigna.
The bottom line: developers pursuing genomic-first basket studies must still break down the evidence by cancer type. If Loxo’s larotrectinib is approved for all cancers with TRK fusions, will insurers selectively reimburse for the cancers that had the best responses? Or will they look at the 55 total patients in the data set and the unorthodox trial design and demand more evidence before paying tens if not hundreds of thousands of dollars per patient? The FDA has become more comfortable with small, unorthodox trials, handing out several cancer drug approvals in recent years that can be rescinded if post-approval studies don’t reinforce the earlier results. Problem is, no so-called conditional approval has been rescinded, even when the drug fails to prove its mettle in a larger population.
It’s possible a genomics-first approach will unearth several more of the right drugs for the precisely right target across multiple cancers. According to the Trinity report, the 37 ongoing basket studies are studying cancers with 16 markers, including well-known cancer-causing mutations like HER2, BRCA, BRAF, and EGFR.
Meanwhile, NCI-Match, the mother of all basket studies, has opened 38 treatment arms, each representing a genetic abnormality with a potential targeted treatment, but it had to expand enrollment last year to fill them up. The goal is 35 people in each arm. If any of the arms shows a signal that the drug is working, the drug’s owner can gather more evidence in a larger trial. Or the National Cancer Institute could pursue a confirmatory study on its own, says Flaherty. But to get clearer signals in NCI-Match and elsewhere, researchers need to accrue more patients and more data. “It’s just the beginning,” says Flaherty. Even if Loxo’s drug is approved for any TRK-fusion cancer, he says, it should be monitored for types of tumors that render the drug ineffective.
“We haven’t unearthed a lot of biomarkers as good as MSI-h,” cautions O’Donnell-Tormey. But the structure of basket studies, if used properly, should make that quest less onerous. Think of it as an architectural problem. Instead of erecting an office building for one company, then tearing it down when the company fails, then repeating the expensive process, basket studies make early research more flexible. “If things fail, you can ask another question,” O’Donnell-Tormey says.
