NCCN Expert Panel Considers Value, Utilization of Molecular Testing in Clinical Practice
Hollywood, FL—The number of molecular tests to assess the potential efficacy of cancer therapies for an individual patient is rapidly increasing, but their value and validity in cancer care remain controversial. An expert panel at the 2011 National Comprehensive Cancer Network (NCCN) Annual Conference discussed how to regulate the tests done, the need for better data to determine their value and cost-effectiveness, as well as new approaches to optimal uses. Lee Newcomer, MD, MHA, Senior Vice President of Oncology Services at UnitedHealthcare, framed the discussion, noting that “evidence” for a good assay comprises 3 points:
- Analytic validity: Do you get the same result 99 times out of 100?
- Clinical validity: Is what you’re measuring going to result in a valid change?
- Clinical utility: Is there something you can do with the information?
Regulatory Standards Lacking for Test Validity
Panelists agreed that greater regulatory standards are needed for molecular tests. “One of the major problems we’re facing is the heterogeneity and variability of these tests,” noted Andrew C. von Eschenbach, MD, former Commissioner of the US Food and Drug Administration (FDA) and Director of the National Cancer Institute, and currently Senior Director for Strategic Initiatives at the Center for Health Transformation. “It’s going to get even more complex as we develop other markers in the area of proteomics and metabolomics. So we’re laying a foundation here, and it’s important that we get it right, because we’re going to have to build on it,” Dr von Eschenbach said.
“Molecular tests are regulated by how they are marketed, not by claims for what the test can do. If a test is marketed as a kit, it is regulated by the FDA. If it tests for certain things—an example is the test that screens the blood supply—it is regulated by the FDA. The FDA would like to regulate on the basis of the claims being made around the test, but they have not stepped in to do that,” said Scott Gottlieb, MD, former FDA Deputy Commissioner and now a fellow at the American Enterprise Institute.
“What has concerned the FDA is that regulation follows, not necessarily the complexity of the test or the claims being made around it, but how the test is being marketed,” Dr Gottlieb said. “So if a test is marketed as a kit, it has to be FDA-approved. If it is being performed as a laboratory service inside a CLIA [Clinical Laboratory Improvement Amendments] laboratory, it is regulated by CLIA and could escape FDA regulation.”
Because of the variability in their performance and minimal regulation, the accuracy of current test results may be questionable, said panel moderator Clifford Goodman, PhD, of the Lewin Group. Dr Newcomer added that 30% of the results from the original trastuzumab registration trial were found to be inaccurate. “This means that 1 of 3 breast cancer patients in the community may be getting the wrong result.” That is, women with a falsepositive HER2 test result “would be getting a year’s worth of therapy that would not be helpful,” Dr Newcomer noted. Those with false-negative results “would be missing a year of very valuable therapy that could save her life.”
Nevertheless, the use of molecular testing in breast cancer has become routine. According to Dr Newcomer, in certain subsets of patients with breast cancer, 94% of the treatment decisions are made based on the risk score of the Oncotype DX molecular test.
Better regulatory oversight of the analytic validity of molecular testing would be expected to correct some of these failings and improve clinical confidence, the panel agreed. Dr von Eschenbach reiterated the need to ensure the consistency of the tests’ claims and results, regardless of where they are performed, as well as allowing the FDA to develop the infrastructure necessary “to bring order out of chaos.”
Dr von Eschenbach stressed the “tremendous amount of resistance on the part of developers, because of the belief that if you placed these tests in the hands of the FDA for regulation, that—by its very nature—would undermine innovation, create significant barriers to progress….I don’t say that those concerns are not well-founded, but I do say that those concerns can easily or systematically be overcome.”
“We need to put it in the hands of the agency that has the scientific infrastructure to be able to make scientific decisions about these tests, as well as be able to judge their clinical utility,” Dr von Eschenbach argued. Dr Gottlieb also supports having the FDA, or CLIA, regulate analytic validity of testing.
According to Dr Gottlieb, there is movement in this direction. Within the next 12 to 24 months, a new regulatory scheme will be applied to molecular testing based on work being done at the FDA and on Capitol Hill at this time.
Nevertheless, “The keystone of the debate is whether or not a regulatory authority should be regulating the clinical validity and the clinical utility,” Dr Gottlieb said. “Determining how important the test result is to the clinical decision that needs to be made is best left to the clinical community.”
Mark G. Kris, MD, Chief of the Thoracic Oncology Service, Memorial Sloan-Kettering Cancer Center, agreed that this is best left to the research community. “It is our job as clinical researchers to provide the data that regulators and payers need to make decisions,” he said.
Toward Personalized Medicine
Dr Goodman asked, “How do we know that these molecular tests work?” Dr Kris, who specializes in treating lung cancer, has been a fervent advocate of molecular testing and has spearheaded much research on its clinical application. He noted 2 instances in which DNA-based testing is firmly driving treatment decisions—testing for the epidermal growth factor receptor (EGFR) in advanced lung cancer and for the KRAS gene in colorectal cancer. In these settings, genetic assays can clearly predict whether a certain targeted agent will be effective in a given patient.
Patients with the KRAS mutation have virtually no chance of responding to the EGFR inhibitors cetuximab (Erbitux) or panitumumab (Vectibix), whereas there is a “decent chance” of tumor shrinkage with these drugs in patents lacking the mutation. In patients with lung cancer, there is a 70% response to treatment with the EGFR inhibitor gefitinib (Iressa) for those with EGFR mutations compared with just 1% in those without the mutation. “Molecular testing is now part of standard guidelines and labeling for these 2 drugs,” Dr Kris said.
But Michael Kolodziej, MD, Chief of Oncology, St. Peter’s Hospital, Albany, NY, said he sees obstacles to utilizing even established molecular tests in his practice. Some tests do not lead to treatments that make a clear difference in patient survival; he experiences difficulties and delays in obtaining tissue specimens needed to perform the molecular tests. “We are aspiring to an era of personalized medicine, but we aren’t there yet,” he said.
At Memorial Sloan-Kettering Cancer Center and a handful of other major cancer centers, oncologists have an easier time with testing. At Sloan- Kettering, patients with non–smallcell lung cancer (NSCLC) routinely undergo a panel of genetic tests, Dr Kris reported, acknowledging that just 1 of the 8 tests in the multiplex assay is backed by phase 3 clinical trial evidence.
Insufficient Evidence, Patient Education
The concern that data supporting molecular and genetic testing are immature was echoed by the panel. “There is evidence, evidence with practical hurdle, and inadequate evidence. EGFR mutational analysis is evidence, but we will not order other tests without good evidence—New England Journal of Medicine level of evidence,” said Dr Kolodziej.
Louis B. Jacques, MD, Director of the Coverage and Analysis Group, Centers for Medicare & Medicaid Services, agreed that clinicians and payers cannot always be certain a given test works. “What we mean by knowing a test works is the same thing that you mean when you look at your 13-yearold and you say, ‘You’re going to be a successful adult.’…The challenge for us is we are looking at this vast array of, say, 8000 adolescents and are being asked to prepay a Harvard education for all of them. And what we don’t know now, because they are adolescents and the evidence about them has not matured, is whether Harvard may be the wrong place for many of them.”
Patient advocate Elizabeth Thompson, who is with the Susan G. Komen for the Cure organization, agreed. “Cancer treatment at NCCN institutions centers on a group of people looking at a wide variety of factors, and these tests are one part of the decision-making module,” she said. “How ever, while the United States is leading the world in terms of what molecular testing means for patient care and outcomes, we still don’t have enough information.”
Ms Thompson added that for many patients, these tests are “confusing and challenging.” Educated patients are aware of these tests and frequently request them, but still have trouble applying them to their specific situations. They often report that their Oncotype DX recurrence risk score is low, yet their oncologist still advocates chemotherapy.
“Patients call us and say, ‘This is what the test said, but this is what my doctor is saying. What should I do?’ Generally, we refer people to a major medical center for a second opinion,” Ms Thompson said.
Dr Newcomer, however, said his data show that 94% of oncologists follow the recommendations for a low-risk score.
“At the end of the day,” Dr Kolodziej said, “I am a cautious optimist. There will prove to be value in some molecular testing, and it will prove to be predictive of response to a specific therapy.” Dr Gottlieb agreed, “If there were no value inherent in these tests, I don’t think they would exist.”
Is There Value for the Dollar?
Is the value of the result obtained with molecular testing worth the high cost of the average assay? This may be hard to determine. The price of testing varies tremendously. For example, EGFR mutation testing costs approximately $800 at Memorial Sloan- Kettering, and this cost also covers testing for an additional 7 gene mutations found in NSCLC, Dr Kris noted.
Other genetic tests, however, are considerably more expensive. “What Dr Kris does not see is the 180-genomic panel that I also get billed for on a routine lung cancer patient,” Dr Newcomer said. “Those tests are 98% unproven.” Dr Newcomer stressed that “a small set of genes can belong together; that makes perfect sense. But we have both ends of the spectrum— people who are being reasonable, and those who are expanding those panels dramatically with no evidence whatsoever. I worry that decisions will be made based on those other 180 tests and they won’t be evidence- based.” Moving forward, he said, “We need to be led by the first precept in medicine: first, do no harm….We are all desperately eager to make these genetic tests work, but we cannot get ahead of ourselves, or we will hurt patients in the interim. We need to organize to get the data. We need to find out what works and doesn’t work, and patiently collect that before we eagerly accept or embrace all of this new technology.”
Attempting to put the issue of cost into perspective, Dr Jacques reminded the audience that “everything about cancer is expensive,” including chemotherapy, hospitalization, and the treatment of serious adverse events.
Other panelists suggested that the benefit will become clear in the longrun. “There is no question that both from a theoretic and an emerging practical perspective, these tests are a methodology for saving money in healthcare,” Dr von Eschenbach noted. The application of molecular testing will reduce unnecessary spending for insufficient or inappropriate therapy, he maintained.
The actual cost of a genetic test, however, is just part of the economic discussion, the discussion made clear. Dr Newcomer revealed some surprising problems with regard to billing. “The current system does not itemize the tests, and the coding system is antiquated,” he said. “It just says ‘genetic test,’ which doesn’t allow us to assess either the upfront costs or the downstream benefits that might result from this kind of testing.”
Dr Newcomer and other third-party payers do not always know where these reimbursement dollars are going. “I have no idea what I am spending, or what I am actually getting, when we reimburse for genetic testing,” Dr Newcomer said. Dr Jacques agreed, “Anything short of clinical utility casts a giant pall of uncertainty over, frankly, whether I ought to be paying for this.”
Dr Gottlieb also noted that logistics issues cited by the clinicians (getting the specimen from the pathology department to the laboratory performing the test) will also become moot, as more integrated entities become the norm. “Anatomical pathologists are consolidating themselves and also purchasing these molecular diagnostic companies,” he pointed out. This should smooth out the process.
Dr von Eschenbach summarized the views expressed by the panel, stating, “These tests are becoming mechanisms for saving money and improving outcomes for our patients. We need to make sure we have the mechanisms in place to make decisions about what is the right treatment, at the right dose, done for the right reason.”