A diverse group, representing the diagnostics development industry, insurance payors, and clinical oncologists, has published a set of recommendations they believe would accelerate the development of clinically useful tests to personalize cancer therapy.
The commentary, which appeared last week in Science Translational Medicine, outlines the structure of what the group calls a 'vicious cycle' in tumor-biomarker research, development, and commercialization, and makes five proposals to solve this problem, including increased — and value-based — reimbursement for molecular diagnostics, reform of the regulatory process for such tests, and increased rigor in the publication and review of biomarker test research.
According to the authors — led by Daniel Hayes, a medical oncologist at the University of Michigan Comprehensive Cancer Center — biomarker-based test development and adoption into clinical use has lagged far behind advances in cancer therapy as a result of a cycle wherein regulatory inconsistencies and poor reimbursement levels reinforce each other, in turn hindering investment in clinical research. This then has resulted in low levels of evidence for the benefits of tests and skepticism about the utility of biomarkers, which feeds back into poor reimbursement and regulatory uncertainties, completing what the group characterizes as a "vicious cycle."
To break this cycle, the commentary included several proposals. For example, Hayes and his co-authors recommend that the US Food and Drug Administration should expand its review process for molecular tests and evaluate the clinical utility of tests, in addition to their analytical validity. The review process for molecular tests should be further reorganized to mirror the review process for therapeutics, they add.
The FDA should oversee all molecular tests, the group says, regardless of whether the test is a kit or developed and performed at a single lab.
Finally, the administration should also recommend that drug trials be accompanied by the creation of an archived biospecimen bank to support biomarker test research.
The group also proposes that reimbursement for biomarker tests should be value-based, and higher for tests with demonstrated clinical utility; investment for test research should be commensurate with that for therapeutics; journals should review tumor-biomarker publications with greater rigor; and professional guidelines bodies should adhere to evidence-based recommendations for test use.
"Our feeling was if we could just get these recommendations out and then work from there to get all these [stakeholders] around the table and talk through on each one of these ideas, I think we could do something," Hayes told PGx Reporter in an interview this week about the commentary. He acknowledged that many of the proposals the group made in the commentary are just "straw men" to fuel such a discussion. However, the authors believe all the sticking points fueling this "vicious cycle" do need to be addressed for real change to take place.
Below is an edited transcript of the interview.
How did you come together with your co-authors to work out these proposals for improving the clinical development of cancer biomarker-based tests for personalized medicine?
I'm a medical oncologist, focused mainly on breast cancer, and I've been involved in the development of tumor biomarkers for 30 years. I've also been involved with the ASCO tumor marker guidelines panel for a long time.
Going through this work, I began wondering why we don't have more [useful] markers, and just as important— why we have markers out there that probably aren’t worth having.
One of my soapboxes has been that a bad biomarker test is as bad as a bad drug. Yet, we don’t insist on the same rigor in basic science and clinical science to demonstrate that these markers really can or should be used to change care as we do [for pharmaceuticals].
These issues together, coupled with increasing interest in so-called personalized medicine in the field, got me to talking with the other people who are authors on this paper and together, and we came up with the concept of this vicious cycle.
Can you explain more about how you see the issues in adoption of clinically useful biomarker tests as being cyclical?
Basically the concept is that it doesn’t matter where you start on this cycle. For instance, you could start with the FDA. The way they regulate biomarker tests – it's much different than with drugs.
To make a long story short, you can sell a biomarker test in this country without having gone to the FDA if you call it a laboratory developed test. And the corollary of that is, you can go to the FDA and get approval for a test that has never been proven to be worth using to take care of patients.
Then you [begin to move through the cycle] because if we make it harder to get biomarker tests through [the FDA], but keep reimbursement the way it is right now — [where] most insurance companies don’t want to pay for a molecular diagnostic test — the entrepreneurs are going to walk away. If there is no payoff, why should they do it?
Then, if we don't have academic rewards and rigor in the way journal editors review these things, we get a bunch of crummy studies published, and then the guidelines panels don't know how to assess the markers because they don't have good data to rely on, and clinicians won't know what is [medically useful], and the cycle continues.
So, each element in this circle influences the next?
Yes. The most important thing is that what we cannot do is just address one of these nodes and not the others.
For example, if we make it harder to get things done but don’t reward entrepreneurs for doing so, nothing will happen. If we reward entrepreneurs for generating lousy data, then we will have bad biomarkers that are being reimbursed for no reason.
All the way around the cycle, everything has to be changed, and at a relatively concurrent time point.
Our thought was, if we just get the dialogue started, then in the long run we can get all the stakeholders in one room talking to each other and we might actually get something done. These recommendations are basically straw men to get this [discussion] started.
Clearly there are a couple of distinct targets here: regulators, test developers, insurers, and the scientific community. To what extent have you engaged these groups to discuss the implementation of some or all of these proposals?
Several of the authors of the commentary represent [the molecular testing] industry … And there are others who gave us advice but didn’t want to be authors [on the paper.]
I have been in contact with the FDA, and they gave some input, and if we are going to move forward, they have to be at the table. Sean Tunis, who used to be head of CMS, is the [commentary's] last author. If we are going to get any traction obviously CMS needs to be at the table, too. And Lee Newcomer, from United Health [was also an author], and they have to be at the table.
Ultimately, the way this is going to happen is that we have everyone sit down and [debate and] negotiate this. It won't be easy, but I think this is the first step.
But some of these suggestions look like they are going to meet with resistance from some stakeholders? For example, the FDA has been talking about extending its oversight to laboratory develop tests for some time, but opponents have argued that forcing LDTs through FDA review would hinder public access to critical tests. What do you think about such arguments?
I'm not a socialist, I'm a capitalist, and I'm also a doctor and I want to have [these tests] out there. But I think in this case [the most beneficial thing] is not to have zero regulation, but to have consistent regulation and a level playing field.
Saying that hospitals shouldn't make LDTs [without] FDA oversight, I don't think that's any different than saying I should not be able to develop a drug in my lab, and as long as I don't cross state lines it's okay to use it.
I can't do that, it's against the law. So, I think if you are developing new assays to help guide patient care, they do need to be evaluated by the FDA, but it has to be a level playing field with good rules, and reimbursement needs to be high enough to make it worthwhile.
It might take [test development] out of the hands of pathologists and that would be a sea change, but I would argue it would probably be a good one.
Of course, there are people who don’t agree, and that’s why we have to be around the table and discuss this.
Why is higher reimbursement for diagnostics so important in the cycle?
We have clinical trial after clinical trial with small, but statistically significant benefits, meaning some patients have benefitted and some have not. Up to now, companies didn't want to know who [those groups] are, because that would limit the market [for their drug.]
But if they develop biomarker tests for which they would get good reimbursement they wouldn’t care [because] whether they get money for the drug or for the test, they still make money.
But now there's no money in tests and we end up overtreating thousands of patients, spending billions of dollars treating people who shouldn’t get the drug in the first place either because they don't need it, or because it's not going to work.
Are there other recommendations you see as particularly controversial in the report?
One of the most controversial recommendations, other than [regulating] LDTs – that’s going to raise some feathers for sure – is suggesting that the FDA should mandate that if a company is running a clinical trial of a new drug, they have to develop a biobank that would be available to society, with the company having some kind of intellectual protection, like a right of first refusal.
Again, it's back to the cycle, because there needs to be support for robust research into the validity and clinical utility of these tests, and that can't happen without the [availability] of samples and data.
Desigining studies to look at clinical utility, not just analytic validity, is good for society and it's good for patients.
One good thing about biomarker tests is that they can be developed using archived specimens if they come from high-level clinical trials, and if those don’t exist, unfortunately you have to do prospective trials.
Overall, are you optimistic about steps being taken in the near future to enact or address the recommendations you've made here?
I don’t have the answers to all this, but my feeling was if we could just get these recommendations out as straw men, and if we could get everyone around the table and work on each one of these, I think we could [make real changes.]
It's not going to happen tomorrow, but I'm optimistic that [this sitting down] will happen.