June 2012, Vol 1, No 2
Personalized Medicine Advances in Melanoma: An Interview With K. Peter Hirth, PhDUncategorized
Plexxikon, a member of the Daiichi Sankyo Group since 2011, works in the structureguided discovery and development of novel small-molecule pharmaceuticals to treat human disease. The company’s lead drug Zelboraf (vemurafenib/PLX4032) was approved by the FDA in August 2011 and is copromoted in the United States by Daiichi Sankyo, Inc. and Genentech. PLX3397, the company’s next oncology candidate, advanced to phase 2 testing in 2011. Plexxikon is developing a portfolio of preclinical- and clinical-stage compounds to address significant unmet medical needs in oncology, as well as in several other therapeutic indications. Plexxikon’s proprietary Scaffold-Based Drug Discovery™ platform integrates multiple state-of-the-art technologies, including structural screening, as a key component that provides a significant competitive advantage over other drug discovery approaches.
Personalized Medicine in Oncology had the opportunity to meet with the CEO of Plexxikon, K. Peter Hirth, PhD, to talk about Plexxikon’s approach to personalized medicine, the landscape of care in melanoma, and the process of biomarker discovery and drug development in niche markets.
PMO We are delighted to talk with you today about Plexxikon’s approach to personalized medicine. We’d like to start by asking how you define personalized medicine, particularly as it relates to the treatment of patients with melanoma.
Dr Hirth Personalized medicine in oncology is not a new concept, but in melanoma it is a very recent development with the breakthrough of the drug discovered by Plexxikon and codeveloped with its partner Roche.
A completely new concept was put forward, and we are now able to dissect the melanoma population into subsets based on molecular drivers. Today, we reclassify and talk about melanomas, not just 1 melanoma.
As a result, we were able to focus on a particular subset of those melanomas, understand the molecular driver, develop the drug, and realize that because this molecular driver is only present in a subset of cancer patients, we would need to come up with a way to identify those patients through the development of a companion diagnostic.
PMO Personalized medicine in action remains sporadic and occurs mainly at well-funded academic medical centers or prompted by physicians who understand the genetic principles behind molecular biomarkers and how to assess them appropriately. How can this situation change so that personalized medicine can be made availiable broadly to patients managed by community oncologists? What are the barriers and opportunities for this to occur?
Dr Hirth Translating and managing the biomarker experiences and the personalized medicine that initially started out in academic centers into the community remains a challenge. However, there are examples where this has been going really well. We are just beginning to understand the barriers to community implementation. For example, the ease of a test is very important. It’s very easy for doctors to deal with binary test readouts; when it’s a simple yes or no. But there are more complicated tests that make for a much more difficult treatment decision. Education is absolutely key to give doctors the tools to interpret the test results.
In addition, when we talk about novel biomarkers and bringing them forward into development, we also have to realize that there are regulatory issues that we have to deal with. The guidance from the FDA at this point is not very clear.
If you look at barriers for the development of biomarkers and their translation into personalized medicine from a business perspective, a biomarker without a drug, without a clinical trial and data, is just a biomarker, and developing that into a full-fledged diagnostic test poses a certain financial risk that people like to measure out as they engage in the development, because it is a chicken/egg situation. I have a biomarker, I might have a drug, but will this drug with this biomarker in combination be successful?
Considering bringing tests to the community oncologist, there is an institution in the US that is somewhat underutilized, and that is the Clinical Laboratory Improvement Amendments of 1988 (CLIA)-certified labs, which really do not fall under the jurisdiction of the FDA but have all the expertise and talent and are already distributed throughout the country. From a distribution perspective, you would actually get much faster penetration into the community if you went that route of involving CLIA-certified labs in the development of these biomarker assays. But this is still a subject of discussion with the FDA, and they’re working on guidance.
PMO Vemurafenib has clearly been a significant advance in the treatment of melanoma, and one can make a case that this agent represents personalized medicine in oncology when used in patients with BRAF mutations. Are efforts by Plexxikon also directed at the diagnostics industry to standardize, innovate, and improve laboratory testing for molecular biomarkers?
Dr Hirth What is the optimum use of identifying the best companion diagnostic? That really depends on the project, and I don’t think there is a one-size-fits-all solution. In the case of Zelboraf and the V600 mutation, it was clear that the genetic level probably would be the most appropriate test. Currently, there is development of further technologies beyond PCR [polymerase chain reaction] looking at high-sensitivity sequencing. So there is going to be a technology change.
At Plexxikon, after looking at the various programs that we are working on, we have decided that basically each program requires a different approach to a companion diagnostic, and therefore we have decided not to engage in developing these tests ourselves but rather go into the marketplace to look for optimum technology available to us.
There is really no single answer to the question of how best to do that. This is our approach, and I think a lot of other companies have gone down that path, and as a result there are a number of diagnostic companies offering their services.
I’m not sure whether this approach will be financially rewarding, because more and more indications are becoming smaller and smaller. The cost for these tests cannot be astronomically high, or nobody would use them. But there is still a lot of work that needs to be dealt with to make this possible in a practical fashion. I believe that we will see platform technologies that will be test run on sequencing platforms. There are going to be platforms like the cell sorting, and individual markers that will be integrated into sets like that. Diagnostic platforms will be developed separately and validated once, so you don’t have to pay the price to develop and validate the entire platform for each new marker, but just the part for the specific new marker for which you want to test.
PMO How is Plexxikon applying its platform to create other oncology compounds beyond vemurafenib?
Dr Hirth The Plexxikon platform that was used to develop Zelboraf has also been utilized for a number of additional programs. Patients eventually develop resistance with cancer treatment and become refractory. The big advantage we have here is our targeted approach; we can go in and look at the resistant tumors and ask what’s changed, what’s different, how do cancers get around. And that now leads to potential points of intervention that we can drug with the technology that we have built at Plexxikon. It will again lead up to matching patients with the appropriate diagnostic test and molecular drivers.
We have 1 compound that we will take forward into development, hopefully early next year, that will be a completely new class of drug in the melanoma space.
In addition, we have PLX3397 that’s in phase 2 development; it represents a very different approach – not by targeting primarily the tumor cell itself but by actually targeting the tumor microenvironment. It’s similar to the concept of angiogenesis inhibition – there is a component of the microenvironment that is critically supporting the growth of tumors.
There will be subsets of tumors that will respond to that. We know that already. It’s a little more tricky and delicate to define the cutoffs – so it’s not as easy as it was for BRAF. I have to remind people that in the BRAF situation it’s really binary, it’s yes/no. Here we’re dealing with more gradation, shades of gray that we have to deal with, so these diagnostic tests will take a little bit longer to define and move forward.
PMO Many of the “success stories” in personalized medicine in oncology have involved collaboration between pharma and a diagnostics company (eg, crizotinib in ALK+ non–small cell lung cancer and vemurafenib in BRAF+ melanoma), and the FDA seems receptive to this sort of collaboration. Are FDA guidelines clear in their requirements for the simultaneous development of a pharmaceutical agent and an accompanying diagnostic test?
Dr Hirth When we talk about personalized medicine and the development from a regulatory perspective, the key features that one needs to consider are in the works at the FDA.
There are attempts to develop draft guidelines, but it’s a case-by-case decision, and I must say that the FDA is very receptive to this companion diagnostic concept because it makes a lot of sense. It is getting those patients with the most need the best and most likely opportunity to benefit from a drug, it shortens development time lines, and it eliminates the cost of unnecessary waste. All of that is very logical and clear.
The question is how it is implemented on a case-by-case basis. The Zelboraf example might not be the standard case because this actually was a more straightforward situation, and less complicated as you think about it from a development perspective, and therefore not necessarily the standard to be measured against in the future.
We will undoubtedly run into situations where we will have to determine how critical new tests are in selecting patients for a particular drug. There’s also the consideration of what regulators want, what payers are willing to pay for, and the types of tests they would like to see done.
There’s a clear example in which payers in colorectal cancer will not pay for EGFR inhibitor drugs: cases with known mutations of KRAS, because studies have shown that the drug is not worthwhile in those patients.
I think the FDA is committed to get this right. They have invited comments. We made proposals, and we also think there are ways to mitigate the initial financial investment by not having a Cadillac rolled out for initial clinical trials but maybe start with a tricycle until we know we really have something that we can develop and then send out around the world. And we think the CLIA labs are a very good setup for that.
PMO Will the collaboration between pharma and diagnostic companies add significant cost to the development of personalized therapy in melanoma?
Dr Hirth The development of a therapeutic together with a companion diagnostic obviously is a change in paradigm, and it does require additional financial commitments, but it also requires a lot more coordination between the 2 companies developing the drug and the diagnostic.
It also requires a lot more communication between agencies at the FDA. The Center for Drug Evaluation and Research is controlling regulating oncology products, while tests are regulated by a very different body, so these groups need to come together and agree on an overall plan.
PMO Despite the fact that the incidence of malignant melanoma has been increasing over the past 30 years, with 76,000 new cases and nearly 10,000 deaths expected in 2012, the discovery of molecular biomarkers associated with this malignancy is still relatively new and, thus far, confined almost exclusively to BRAF (although data are now being accumulated for the association of c-KIT and NRAS mutations with melanoma). Why do you think the discovery process of molecular biomarkers in melanoma has been so slow?
Dr Hirth Why has melanoma been such a graveyard in drug development over the last year? Why did we not have any drugs before Yervoy and Zelboraf? Does this have to do with the lack of biomarkers? No, not really. Plenty of biomarkers were identified. There was interleukin- 1a. There was the interferon. There were all kinds of interleukin-2 stories that were available. But we never came to the point where anything was tested to form hypotheses for subtypes. It was all looked at in a traditional drug development way in a heterogeneous patient population.
The other downside of melanoma is that it’s a disease that is so aggressive. That may be why other cancers that have a longer development time line before they become fatal have been better from a drug development perspective.
Generally, the capability to generate potential biomarkers over the last few years has gone up exponentially compared with the past, primarily because of high-throughput sequencing, and I see throughput sequencing getting cheaper and cheaper, so many academic institutions are now collecting data. My concern here is who’s sifting through all the data, who’s making sense of it. And then from the data we need to build hypotheses to ask questions. Well, are these tests epiphenomenal? Are these passenger mutations or do they actually drive indications?
Once we have identified driver mutations, if they’re druggable, we know what to do in terms of drug development. But it’s not going to be something that will be changing overnight. I know there are some people who have this expectation with sequencing. The field will be changed in 2 years. I like to manage expectations; from the time we identify a potential target to having something on the market is still many years away.
We have shown that with a companion diagnostic we can accelerate the drug development process quite substantially, but it will still be a number of years before we see the benefit of these new technologies coming through.
PMO It seems that education of providers, pharmacists, payers, and patients is vital in achieving personalized medicine in oncology. What efforts is Plexxikon making in educating these stakeholders?
Dr Hirth Education about these new approaches to personalized medicine obviously involves a variety of different parameters, functions, clinicians, payers, insurance, and advocacies. We have initially taken the road that we would start out in a scientific community, because this is the culture that we live in, where new concepts are first tested in an academic environment before they trickle down into a community setting.
But it has been going really well – especially in melanoma, where patients have been so frustrated – because there wasn’t anything else new on the horizon; the advocacy groups were looking for new events and new approaches that could change the outcomes.
I would imagine that payers will have a huge interest in embracing this concept, because it will reduce wasted drug on patients who are unlikely to benefit, and I cannot see how that is not a very important argument for payers.
Now, where they would go with testing, whether they would follow the FDA, or whether they would like their own testing labs, this is still a subject of discussion, and I don’t know how this will come out.
PMO What do you foresee as the future of personalized medicine in oncology when sequencing the entire human genome will cost less than $1000 in the next few years? How will this impact drug development for melanoma and other cancers?
Dr Hirth We’re seeing huge progress in sequencing efforts, being able to sequence entire genomes, genomes of cancers, and at a price that is falling to a level that makes it realistic to apply it to a really large population.
How will that change the way we treat patients? First of all, clearly I’m looking forward to being able to better identify populations within, let’s say, melanoma tumors or breast tumors that share certain characteristics. We can learn what the most likely molecular drivers are, and if they are druggable. We will have a huge opportunity to come up with new approaches to treatment.
So what I think will happen is that we’ll get better treatments, we will get safer treatments, but the treatments will be more selective. There will be more niche drugs. There will be smaller groups of patients, and that has some challenges in terms of conducting clinical trials and in the necessary investment. To develop a drug is a very expensive exercise that is justified mostly because there’s an anticipated patient population that you can treat and subsequently make money. But as these populations get smaller and smaller, we have to ask ourselves to identify the break-even point.
I think we need to have changes in the regulatory environment and in the reimbursement environment. We need to reconsider what our end points need to be, how much time and money we can afford to spend in some of these smaller niched indications.
We already see this at Plexxikon, where we start smaller trials – 10, 20 patients.
It is more important for us to show early on in development that we hit the target, that we can establish the concept of how the drug is working. Even if that is a small indication, but then later on, once we have understood that, we can bring it to the broader audience.
And I think what we do with Zelboraf is very similar. We have shown in melanoma how that works. This is now being tested in other tumors as well. We don’t know the outcomes yet, but I’m positively optimistic as always, otherwise I wouldn’t be in this business.
PMO What will you offer melanoma patients without a biomarker that directs a targeted therapy, especially if this impacts a significant percentage of patients?
Dr Hirth When we look at the melanomas and what biomarkers are present, we still have about a 40% proportion of the population where there is really no clear biomarker identified that can lend itself to developing a targeted drug. This is unfortunate, because there is not much available now for those patients other than dacarbazine and Yervoy.
I would also like to say that even though we now have these subpopulations based on V600 BRAF, this is not a complete homogeneous population either. It’s enriched, but it’s not black and white, so there are patients who respond strongly and patients who have less of a response. So clearly there are further fractionations that will happen. In the future, I do see a more niched approach to therapy.
The other thing I would expect to see is that we will have a much more rational approach to combining Zel – boraf, for instance, with another drug, and we can start calling these companion therapeutics. In the old days, everything was combined with everything available, but there was very little logic and science behind it. We see that changing. We already have very beautiful examples in some of the BRAF-driven tumors, where the most appropriate companion therapeutic is basically already on the market and should be tested.
So this is a change there as well. It will be much more rational. And again, it leads to better development, faster time lines to bring new products and benefit to patients in the clinic.
But I think it also requires a rethinking in the industry that has consequences in organizational forms. A patient population consisting of 2000 patients doesn’t require a 20,000 people sales force, for example. And so we will see dramatic changes in the oncology business as a result of the science driving the search for more niched opportunities becoming better through the use of companion diagnostics approaches.
PMO Will it still make sense for Plexxikon to develop agents that impact only 5% to 10% of patients?
Dr Hirth What is the patient population in any given disease that is attractive for development? This is a very difficult question to answer. Personally I believe that whenever we can have a very dramatic impact on patients who are in need of help, it’s worthwhile doing. We have some extreme examples with Genzyme, for example, which has developed really rare indications.
We know that targeted therapies that have a high impact usually are priced better. The value proposition is very different, and that is reflected in the investment. It can hopefully be developed faster so maybe it doesn’t require the $800 million that it typically would take.
So I think there are opportunities in the biotech space to embrace concepts of developing drugs more quickly and more financially efficiently to make it a worthwhile proposition. I think it’s worthwhile any time when we can help people; it’s a worthwhile proposition for me.
PMO Thank you so much for taking the time to talk with us today.
Advocating for biomarker-driven “smart” trials, making the case for patient selection for specific drugs, and the concept of rational combinations as potential targeted therapies in order to improve the efficacy of clinical trials were topics of an insightful presentation delivered by Dr Aleksandar Sekulic on behalf of Dr Pat LoRusso [ Read More ]
In a preliminary phase 1 study, the targeted drug crizotinib (Xalkori) stabilized disease progression and, in some children, eradicated cancer cells in relapsed, refractory, aggressive childhood cancers that are likely to harbor genetic abnormalities in the ALK gene: anaplastic large cell lymphoma (ALCL), inflammatory myofibroblastic tumors (IMT), and aggressive neuroblastoma. [ Read More ]