For people with advanced cancer who are running out of options, many cancer centers now offer this hope: Have your tumor’s genome sequenced, and doctors will match you with a drug that targets its weak spot. But this booming area of cancer treatment has critics, who say its promise has been oversold. Last week, two prominent voices in the field faced off in a sometimes-tense debate on what’s often called precision oncology at the annual meeting of the American Association for Cancer Research (AACR) in Chicago, Illinois. Their dispute threw a splash of cold water on a meeting packed with sessions on genome-based cancer treatments.
On one side was David Hyman, a 30-something oncologist at Memorial Sloan Kettering Cancer Center in New York City and a leader of clinical studies testing gene-targeted cancer drugs. Genomics won’t help most cancer patients at this point, he acknowledged, but many do clearly benefit: “I think this is certainly not hype,” Hyman said.
Countering Hyman was another young oncologist, Vinay Prasad of Oregon Health & Science University in Portland, a prolific author of provocative journal articles and outspoken Twitter commentator with more than 17,000 followers. Prasad dispenses sharp criticism about various issues in medicine, from drug prices to conflicts of interest. In the case of precision oncology, he argues that far fewer patients will benefit than proponents suggest. “When you look at all of the data, it’s a sobering picture,” he said.
Targeted drugs for cancer date back to Herceptin, approved for breast cancer in 1998, and the 2001 approval of the leukemia drug Gleevec, both of which have saved many lives. So far, the U.S. Food and Drug Administration (FDA) has approved 31 targeted therapies for various cancers. Like Gleevec, most work by blocking mutant cancer-driving proteins, although one type harnesses the immune system to fight tumors.
Hyman stressed that growing numbers of patients qualify for the treatments. At Sloan Kettering, which has tested more than 25,000 patients’ tumors, 15% can be matched with one of the FDA-approved drugs, and 10% with a drug in clinical trials, he said. Another 10% to 15% or so have a DNA change that matches a potential drug tested in animals. Other cancer centers are reporting similar match rates.
Trials known as “basket” studies are likely to expand the patient pool by showing that these drugs can work against multiple tumor types that carry the same key DNA changes. In February, for example, Hyman and co-authors reported in The New England Journal of Medicine that a drug aimed at genes known as tropomyosin receptor kinase (TRK) fusions works for many tumor types.
FDA endorsed the first “tissue-agnostic” cancer treatment last year—an immunotherapy drug that is now approved for any advanced solid tumor with flaws in genes that repair DNA. More are expected soon. And although TRK fusions and the DNA repair mutations occur in only a small fraction of patients with a particular cancer type, when tallied across cancers, such drugs can help many patients. As the list of targeted drugs grows, it makes sense to test tumors with genome-wide assays, Hyman said. “If we don’t test people broadly, we will miss patients who have alterations for which there is now an approved therapy.”
But Prasad sees a glass half-empty when he combines data on how many patients respond to genome-based drugs. He agrees with Hyman that the pool of potential patients is growing. In his latest analysis, published online last week in JAMA Oncology, he and his co-authors found that in 2018, 15.4% of 610,000 U.S. patients with metastatic cancer were eligible for an FDA-approved, genome-guided drug. But he also found that because the drugs shrink tumors in only some eligible patients, just 6.6% likely benefited. And many patients relapse after a couple of years on the drugs.
Even the new drugs being tested in basket trials have just a 22% response rate overall, which is no better than chemotherapy, Prasad argued in an unpublished analysis he presented. At the current slow, steady rate of new FDA drug approvals, it would take more than 200 years for all patients to be helped, he said.
He’s also troubled by a recent federal decision to allow Medicare to cover a genome sequencing test that analyzes tumors for 324 cancer genes. Prasad thinks these test results could lead some oncologists to prescribe a “bad drug” that matches a mutation but hasn’t been proven to work for that cancer. He argues that the testmaker should have first conducted a clinical trial comparing how long patients live after receiving a drug suggested by a genome test to patients whose treatment is based only on conventional pathological analysis of their tumor. “You could answer the question in just 1 year,” he said during his AACR talk.
A testy discussion followed. Moderator José Baselga of Sloan Kettering and Hyman criticized Prasad’s analysis of the basket trials, saying these exploratory studies are not meant to give firm data on response rates. Baselga also suggested oncologists would be sued if they didn’t order genomic testing for lung cancer patients, who often can benefit from a targeted drug. Hyman added that it’s “paternalistic” to say physicians can’t handle tumor genome data.
The two young cancer researchers found a few points of agreement, however. Prasad praised the basket trials that Hyman helps lead for generating rigorous evidence supporting some targeted drugs: “We actually think similarly,” he said. And Hyman conceded that “there’s a lot of work [still] to be done.”