Cancer mapping shows limited benefit
NEW HAVEN — In this age of personalized medicine, sequencing the genes in a patient’s tumor can reveal mutations that may be treatable with targeted medications.
But a new Yale School of Medicine- led study published Aug. 7 in the Journal of the American Medical Association shows that testing for many such mutations does not improve the chances of a patient’s survival.
Data was analyzed from 5,688 patients with nonsmall cell lung cancer who were treated in a community cancer clinic rather than a large research institution. About 15 percent received broad- based sequencing of the tumor’s genome; the rest were tested only for two mutations, known as EGFR and ALK, for which medications are available.
Incorporating variables into the analysis, the researchers found that mortality rates after 12 months were 41.1 percent for those who had the broad- based sequencing and 44.4 percent for those who just had the tests for EGFR and ALK.
“I think our ability to sequence our genes in lung cancer tumors has really outpaced our ability to get patients the drugs that they need when they find a genetic mutation,” said Dr. Carolyn Presley, lead author of the paper and a geriatric oncology specialist at the Ohio State University Comprehensive Cancer Center, who worked at Yale during the research. “In this study, we did not see a survival difference between testing two genes vs. testing 30 or more genes.”
That is largely because drugs have yet to be developed to switch off those mutations, which drive the growth of the cancer, Presley said.
‘ Baseline’ for treatment
Other issues raised by the findings include the need for faster approval of new drugs by the U. S. Food and Drug Administration, for health insurers to pay for experimental drugs and the high cost to patients, Presley said.
“It gives us a baseline,” said Dr. Roy Herbst, a study co- author and chief of medical oncology at the Yale Cancer Center and Smilow Cancer Hospital. “It tells us, now we know, despite all the gene testing, the survival benefits are not being seen because we have not matched the patient with the right drug.”
Among the lessons, according to Herbst, is the importance of clinical studies to test medications to treat more genetic mutations. He said there are about 50 such trials being conducted at Yale to find drugs that will act on more mutations.
“Without large clinical trials and data, you’re treating each patient separately,” Herbst said.
In order to do that, “once every two weeks here at Yale, we have a precision medical tumor board” that discusses how best to match patients with the drugs available, he said.
The study also found that few of those whose tumors were tested for a wide range of mutations received treatments targeted to their mutations. “A lot of patients who got tested for multiple genes, we would have expected them to get this targeted treatment,” Presley said. However, “there was not an additional benefit over testing for EGFR or ALK.”
High costs
Presley said that in addition to the lack of drugs developed for additional genetic mutations, other medications weren’t given because health insurers would not approve them, or because they weren’t approved by the FDA or because the co- payments were too expensive for the patients to afford.
She said there were “other mutations that we would have expected patients to get targeted treatments for but they did not”: ROS1, MET and BRAF. Medications for those “were not FDA- approved at the time of this study,” she said.
“If we repeated that study basically between 2016 and now, maybe we would see improved outcomes because the field of lung cancer is moving that fast,” Presley said.
She said it “is expensive to do the test itself but the cost of the test really pales in comparison to the cost of drugs.”
Presley, who was recruited to Ohio State to “build a comprehensive cancer and aging program,” brought up the public policy issues raised by developing new drugs targeted to individual patients’ cancers. “We have the right to try but there’s not a right to be able to afford your medications and I really think that’s a major issue here,” she said. “We can’t expect our patients with advanced lung cancer to pay thousands of dollars a month for their medications because they don’t want to bankrupt their families.”
When drugs are not approved for a particular disease, for “every patient we have to ask the drug company to reduce cost or we have to ask them to help the patient involved,” Presley said. “We have whole teams of people to fill out the paperwork,” she said. “It’s incredibly cumbersome. It can delay [ treatment] by weeks.”
Assessing helpfulness
Dr. Cary Gross, director of Yale’s Cancer Outcomes, Public Policy and Effectiveness Research Center, and senior author of the JAMA paper, said, gene sequencing now “can test for dozens or … even 300 mutations of the gene pattern at the same time,” Gross said. In fact, according to Herbst, since 2016, treatments have been developed for mutations in genes such as ROS1, BRAF, RET, NTRK, HER2 and PDL1. The mutations are not necessarily found in smokers either. The EGFR mutation is found in more nonsmokers than smokers, according to Presley.
While it’s important to continue testing for a wide variety of genetic mutations, Gross said, “As we’re rolling out these new genetic tests, we should be looking to see if they’re helping people. … We need to be really carefully assessing whether these tests are helpful.”
He said the JAMA study focused on patients at community cancer centers. “We’re not talking about big research institutions like Yale. … Frankly, at institutions like ours we should be doing these big clinical tests. … The question we wanted to address is, can you get these multiple genesequencing tests performed at community practices. We wanted to see how often
[ do] the results of these tests guide treatment decisions.
“In some of these settings, if the oncology practice is not using this data as precision medicine … with very little data out there a patient could be getting a treatment for a gene abnormality when we don’t know if it is helpful or not. For patients who are getting these genetic tests, they should make sure they’re going to a center where the doctors and the team are very familiar with how to interpret these genetic tests.”
Despite the lack of improvement in survival among those whose tumors were subjected to broader gene- sequencing, Herbst said, “I’m in favor of profiling. I think it’s important to know as much as we can about the enemy.”
Both Herbst and Gross called the study “a call to arms.” Gross said that’s true “on two fronts”: “One is to find new treatments. … We should also be evaluating new testing approaches. … As we’re rolling out these new genetic tests we should be looking to see if they’re helping people.”