The European Personalized Medicine Association held a webinar last month dissecting some of the many challenges of expanding personalized medicine markets in Europe.
As part of the presentation, Nick Crabb, associate director of the Diagnostics Assessment Program at the UK's National Institute for Health and Clinical Excellence, discussed some of the issues the agency expects to encounter when evaluating heath technology as the pool of drug and diagnostic personalized medicine products grows.
Crabb raised several hypothetical scenarios in his presentation, highlighting challenges and possible solutions for evaluating cost-effectiveness of companion diagnostic products, including relatively weak evidence links between diagnostics and patient outcome, and differing strategies necessary for evaluating co-developed drug-diagnostic pairs and tests developed independently of drugs.
Crabb told PGx Reporter that some of scenarios he discussed in his presentation have been used as a framework for discussing companion diagnostics issues within NICE, but that they do not reflect official NICE policy for evaluating drugs with companion diagnostics, which are expected to develop over time.
In an interview with PGx Reporter, Crabb discussed the institute's role in evaluating diagnostics for UK healthcare as part of what it expects to be a growing presence of personalized medicine products in NICE's purview.
Below is an edited transcript of the interview.
Can you tell me a little bit about the development of the Diagnostic Assessment Program and any new efforts at NICE to assess cost effectiveness and clinical effectiveness of drug and diagnostic personalized medicine products? Are you adding any new resources or capabilities to keep up with this rapidly evolving area in drug development?
The new Diagnostic Assessment Program was set up by NICE in 2009-2010 to evaluate medical diagnostic technologies. The aim of the program is really supporting the NHS in adopting clinically and cost-effective technologies more rapidly and consistently. We're not really set up as a gatekeeper — so we're not looking at every diagnostic technology. We're looking at a relatively small number of technologies but ones that have the potential to either produce real patient outcome benefits, or system benefits in the form of cost savings.
The program is fully up and running and we are currently working on five topics [see here for further information — Ed]. The new program represents a major increase in NICE resources for the assessment of diagnostic technologies, some of which may well be applied to the consideration of companion diagnostics. Detailed arrangements for the assessment of drug/diagnostic products may develop over time, but we consider that the long-established Technology Appraisal program, together with the new DAP, provide a firm foundation for the evaluation of these products
How do you decide which products or technologies to evaluate?
NICE’s Medical Technologies Advisory Committee selects medical technologies for evaluation, including diagnostic technologies. MTAC has defined criteria for selecting topics and routing them for evaluation by NICE. Suitable diagnostic technologies are evaluated by DAP. Topics are actually nominated by sponsors and the sponsors are typically the manufacturers. If a manufacturer has a technology that they think can really help patients in the form of better patient outcomes or it can save the healthcare system a lot of money, they can inform NICE. Those notifications are reviewed by the medical technology advisory committee, and then appropriate topics are selected and routed to the most appropriate NICE program.
So, that's really how it works, and the program was set up that way to encourage industry to inform NICE of innovative technologies that could offer benefits to patients.
But in the diagnostics program, we have an interesting feature whereby, although the sponsor notification process is likely to result in the notification of specific proprietary products, within the DAP process we can expand the scope to include consideration of alternative related technologies. So even though we were notified of one particular topic and had that referred to us, we can do searches around it. And if there are quite a few technologies working in the same space we can bring them all into the assessment.
Once you have a topic on your plate, what is the process the program goes through in assessing that product?
We're looking at clinical effectiveness and cost effectiveness.
One of the features of our committee that is quite important is that, as well as having standing committee members, we also recruit specialist members for each topic. That allows us to fine-tune the capability of the decision-making body to each individual topic. We think that's very, very important to diagnostics because the diagnostic technology itself isn't providing the health outcome benefit. The health outcome benefit is delivered by all the clinical decisions that have been informed by that diagnostic. It means we need to model patient benefits through the post-diagnostic care pathway, and in order to do that it is very important to make appropriate assumptions in that modeling. To make sure that we can do that, we need a very high level of clinical expertise. That's one of the key reasons we have specialist committee members in the composition of our committee.
As you anticipate assessing the cost-effectiveness of more Rx/Dx products, do you see the threshold for those being different from a drug that isn't targeted to a genetically defined subpopulation?
In terms of drug and companion diagnostic combinations, the detailed arrangements for assessing those will probably develop over time, but really the programs that are likely to be involved will be the diagnostic program we've been talking about, but also the technology appraisal program, which looks at pharmaceuticals. I don’t think all companion diagnostic products will necessarily end up in the DAP. And so it really will be a combination of the two programs.
The arrangements for companion diagnostics are likely to evolve, but right now we use the technology appraisal process and the diagnostic assessment process. We don't currently have plans for any different thresholds for companion diagnostic products. We'd be expecting to apply the same sort of thresholds we do in the two programs.
Can you discuss any PGx applications that currently being assessed by NICE?
So far there are relatively few examples, but … this whole area is getting much more important, and as a consequence of that, we expect the number of pharmaceuticals with a diagnostic dimension to them to increase.
One example of a NICE appraisal that has involved a diagnostic is trastuzumab as a treatment option for women with early-stage HER2-positive breast cancer. In that case, the economic modeling undertaken did include the costs of HER2 testing, and also, within that appraisal, the diagnostic accuracy data was considered and obviously because the clinical outcomes had been informed by the HER2 testing, those outcomes we used in cost-effectiveness analysis were the outcomes that resulted from that combination of the pharmaceutical and the diagnostic.
In your discussions and reviews, what do you identify as the challenges encountered by sponsors developing genomically targeted personalized medicine products in meeting your efficacy and cost effectiveness thresholds?
It’s a tricky question. Obviously sponsor organizations have all sorts of challenges around companion diagnostics. Developing diagnostics and developing drugs are both incredibly difficult, onerous, expensive things to do. The overall costs associated with developing drugs and diagnostics are huge challenges, but I can't think of any specific challenges that are associated with meeting NICE thresholds.
There are differences often between the drugs that NICE will greenlight versus what the US Food and Drug Administration will approve. For example, in your review of AstraZeneca's Iressa in the EGFR-mutated population, the drug developer didn't meet overall survival endpoints but met progression-free survival endpoints. This wasn't sufficient for the FDA and AZ ceased trying to gain full approval in the US. Can you account for this difference in regulatory stance between US and the UK? In your opinion is the US being too rigid in its reviews of PGx products?
We can't really comment on that, because there is a big difference between the health technology assessment and the regulation in the UK. NICE is the health technology evaluation agency, it's not a regulator. So our starting point will be before we evaluate any sort of diagnostic, it needs to be either legally available in the UK, or well on track to becoming legally available. If something has been approved, or is well on its way to being approved, we will start our assessment, and the appraisal will only be issued once the drug has received its marketing authorization from the regulator, which in the UK is the [Medicines and Healthcare Products Regulatory Agency].
With a lot of pharmaceutical assessments, NICE aims to use its relatively fast-track process, called the single technology appraisal, and aims to issue guidance around the same time that the drug will get regulatory approval. If for any reason the drug does not get regulatory approval, then of course that guidance will not be issued.
Is it possible that a product will be approved in terms of regulation, but won't be recommended by NICE?
That can happen. NICE recommends far more products than it doesn't recommend, but the scenario does happen, yes.
What about a situation where, although a drug and test have been developed together, your review finds a test that has better clinical utility and cost considerations than the co-developed test? What do you do in that scenario?
This is an important question. Part of the answer is noting the difference between assessment and regulation again. NICE can only evaluate pharmaceuticals within their licensed indications. What I don't think is yet clear is how the general regulatory frameworks are going to evolve for pharmaceuticals and companion diagnostic combinations.
If, for example, a companion diagnostic and a pharmaceutical were co-developed and the marketing authorization or the label requirements, however you want to say it, stipulated use of the specific co-developed test, then it wouldn't be appropriate for NICE to consider alternative tests. I'm not sure if the regulatory framework is likely to evolve that way or not.
Now, if on the other hand, the marketing authorization for a pharmaceutical required a diagnostic, but there were multiple test options that would still meet the requirements of the drug's marketing authorization, then it's possible that a test that hadn't been co-developed was actually the better test, in which case the situation could happen.
Is this something you've seen thus far?
If we go back to the example I gave earlier of HER2 testing, if you actually look at the requirements of the marketing authorization, it does talk about using validated methods for HER2 testing, but doesn't specify in great detail a specific test. So theoretically somebody could come and notify a HER2 test to the DAP, for example.
If they were to do that, and the Medical Technology Advisory Committee selected it and routed it to the DAP, then theoretically we could look at the various options for HER2 testing. But that hasn't happened so far.
The program is very new. And within the work that was done for the trastuzumab technology appraisal, they did look at the diagnostic accuracy of the data of the test that was actually used in the clinical trials, but it didn't look at any alternatives.
What are you doing to align your personalized medicine regulations with others in Europe, for example the P-Medicine effort by the European Community's 7th Framework Program? What is EPEMED's role in helping bring consensus throughout different European regulatory bodies around personalized medicine?
NICE is engaged in a number of initiatives aiming to support the development of personalized medicine. A major one is the UK Technology Strategy Board. They have what they call an innovation platform on stratified medicine and NICE is a partner organization on that.
We are also involved with an initiative in Europe called EUnetHA [European network for Health Technology Assessment], which is really a collaboration on health technology assessment, focused broadly on the assessment of all types of technologies, pharmaceuticals, interventional procedures, diagnostics, medical devices. That's much broader, there's not a specific personalized medicine component, but it’s a broad general collaboration.