BUSINESS DAY

Doctors Seek Test for Deploying New Life-Extending Cancer Drugs

By 
Dr. Jedd D. Wolchok, chief of the melanoma and immunotherapeutics service at the Memorial Sloan Kettering Cancer Center. Memorial Sloan Kettering Cancer Center 
CHICAGO — New drugs that boost the immune system’s ability to fight tumors may be one of the greatest medical advances in years, cancerdoctors say, pulling some patients from death’s door and keeping them in remission for years.
But the truth is that this happens for only a minority of patients. Now, doctors say, there is a new imperative to develop a test that will identify in advance which patients will benefit, sparing others the cost and possible side effects.
The drugs currently cost about $150,000 a year per patient — even more for higher doses used in some cases — and the health system is eventually expected to spend billions or even tens of billions of dollars on the drugs each year.
“We don’t want to give these to 100 percent of the patients if only 59 percent or 20 percent will benefit,” said Dr. David R. Gandara, a professor and lung cancer specialist at the University of California, Davis. Being able to test for a biomarker that could predict the drugs’ efficacy “would make this new class of drugs easier on the wallet, the national health wallet,” he said.
But developing such a test has proved tricky so far, for ethical as well as scientific reasons. Some doctors said it would be unfair to withhold the new drugs from patients based on a test if there were still even a slight chance that the drugs would help.
“We don’t want to be wrong, because these medicines have an effect that, in some cases, is durable for years,” said Dr. Jedd D. Wolchok, chief of the melanoma and immunotherapeutics service at the Memorial Sloan Kettering Cancer Center. “We don’t want to have an imperfect biomarker.”
The need for such biomarkers is illustrated in a study led by Dr. Wolchok, which he presented on Sunday in Chicago at the annual meeting of the American Society of Clinical Oncology. The study is published online by the New England Journal of Medicine.
The 945-patient study shows that the combination of two immune-boosting drugs from Bristol-Myers Squibb — Opdivo and Yervoy — is more effective than either drug alone in treating advanced melanoma. Patients treated with both drugs went a median of 11.5 months before their disease worsened, a longer reprieve than the 6.9 months for those who received only Opdivo and 2.9 months for those who took Yervoy.
But the combination also caused serious side effects like diarrhea and colitis in 55 percent of patients, compared with only 16.3 percent for Opdivo alone and 27.3 percent for Yervoy alone.
Dr. Antoni Ribas, a melanoma specialist at the University of California, Los Angeles, who was not involved in the study, said Opdivo alone might be just as good as the combination for many patients, with far fewer side effects, but that a biomarker test was needed.
“The combination is outstanding, but we have to figure out who needs the combination as opposed to the single agent,” he said.
The main test being explored is for PD-L1, a protein produced by cancer cells that, in effect, orders the immune system to stand down and not attack.
The Merck drug Keytruda, Opdivo and other similar treatments work by keeping this “stand down” order from being received by the immune cells. So it makes sense that the drugs work best against tumors that are issuing such an order and that they may not work at all against tumors that are not issuing the order.
Studies by Bristol-Myers and Merck, as well as Roche, which is also developing such a drug, have shown that there was a much greater success rate using the drugs to treat tumors that were positive for PD-L1.
Still, at least a small number of patients whose tumors do not produce meaningful amounts of PD-L1 also seem to benefit from these drugs. So some doctors say it is wrong to withhold the drugs from patients whose tumors test negative for PD-L1.
In the melanoma study, patients whose tumors were positive for PD-L1 did as well on Opdivo alone as with the combination, as measured by the delay before their cancer worsened. One implication might be that those patients should get only Opdivo, while others should get the combination.
But Dr. Michael B. Atkins, deputy director of the Georgetown Lombardi Comprehensive Cancer Center in Washington, said that even for PD-L1-positive tumors, the combination was better at shrinking the abnormalities.
“The biomarker isn’t good enough to make any decisions on it,” said Dr. Atkins, who was not involved in the study.
PD-L1 is not the only possible biomarker. Scientists are finding that the drugs work best against tumors with lots of mutations. Researchers reported on Friday that a genetic signature could identify a small subset of patients with colorectal and other types of cancer who would be likely to benefit from Keytruda.
Dr. Ribas and colleagues suggest examining tumor samples to see if immune cells are present. The drugs appear to work best when immune cells are already in or near the tumor, ready to attack when the “stand down” order is lifted by a drug. If the immune cells are not present, then merely lifting the order may not be enough.
Merck is working with a diagnostic company, NanoString Technologies, to develop a test that measures activity levels in genes associated with immune response.
A downside for drug companies is that a test can narrow the market for a drug.
Shares of Bristol-Myers fell nearly 7 percent on Friday based on what would seem to be positive clinical trial results showing that Opdivo could prolong the lives of patients with the most common form of lung cancer.
But there was a big survival difference in patients with PD-L1-positive tumors and patients whose tumors test negative for the protein. For those with PD-L1-negative tumors, there was no real difference between Opdivo and the generic chemotherapy drug docetaxel. This information dashed investors’ hopes that Opdivo might be used by all patients with that form of lung cancer.
Opdivo did cause fewer side effects than docetaxel, but insurers might not be willing to pay so much more for that reason alone.
Docetaxel costs $6,000 for six cycles of treatment; Opdivo used for the same length of time costs about $60,000, said Dr. Patrick W. Cobb, an oncologist in Billings, Mont.
“The cost of treating these patients will be far higher than in the past,” Dr. Cobb said on a webinar sponsored by Kantar Health, a consulting firm. “We really need a way of determining which patients are likely to benefit from these agents.”