Compound- to Target-Centric: A New View of Drug Discovery with H3 Biomedicine CEO Markus Warmuth
October 29th, 2012
Posted by Shane Climie, Ph.D.
The cost and risk associated with biopharmaceutical R&D is enormous and the rewards are increasingly elusive. For many young firms, the “valley of death” is a very real obstacle. What can startups (and larger companies) do? In the first blog post based on my interview with H3 Biomedicine CEO Markus Warmuth, we discussed H3’s strategy to propel innovative therapies into the market using a tighter time frame based upon a new research and clinical paradigm that is almost the reverse of traditional biopharmaceutical development. In this segment, Warmuth shares with us some specifics about H3 Biomedicine’s unique approach to drug development.
Q: What is the current approach to drug discovery at H3?
Warmuth: It is very target centric. Our discovery and development is focused heavily on target identification and validation. When systematic gene expression studies started to come out, it came with a lot of hope but then failed the industry. We found out that a change in the expression of a gene doesn’t necessarily mean a change or impact on the course of a disease. Instead, gene mutations have proven very powerful in predicting disease relevance. And, through public efforts like The Cancer Genome Atlas (TCGA) now we have more information at hand to draw conclusions quickly about the specific consequences of mutations. In some cases we find that alternative splicing can provide some insights, and this is an area we are very focused on. However, we have observed that mutations – and those are the majority – can inactivate or change the function of a gene or its product such that the mutated gene might no longer be a viable drug target. Under such circumstances we typically need to find targets that lie downstream or are synthetic lethal to the target – an area we are exploring very heavily.
Q: What is the best method for target validation in drug discovery?
Warmuth: There’s no simple answer. You have to go gene by gene and ask, “What are the key scientific questions that would support the nomination of a given gene or its product as the target for a discovery program?” The right method or technology is the one that gets you to the finish line with the most convincing data. Today, you can’t bank on one specific technology. We’re trying to diversify our methods to find the right answers related to different kinds of targets.
Q: How are you different from “big pharma”?
Warmuth: I’m not claiming that our approach is better, but it’s unique—it’s completely focused on our strategy. In a traditional development paradigm, there are a lot of clinical failures, especially in Phase 2, because the discovery process is very compound-centric. The only time most researchers in drug discovery think about the target is at the beginning of a project, prior to a screen. From then the process and milestones are very compound (or NCE) centric. They do a screen to find hits, they take the “hit” to a lead, keep optimizing the lead candidates, select, and continue to optimize the compound until they’ve arrived at a compound ready to put in the clinic. In this process, many times scientists forget to ask whether the original hypothesis around the target still holds up, because they are rewarded for bringing a compound forward. So it’s all compound-centric. We’re trying to support a target-centric approach. We challenge the target and our hypothesis at every step of the preclinical process, with different methods and compounds. Because of that, we will have by the time we enter the clinic so much confidence in the target that, we’re not trying to just “see what we get.”
Q: What is a druggable target?
Warmuth: That is a target that is amenable, and responsive to current drug discovery approaches. Some targets are considered to be “undruggable,” if their activity cannot be modulated using compounds that are found in typical libraries of drug – like molecules. But with the right library, you might be able to “drug” the target. There are two important considerations here:
- Chemical space; some compound libraries are geared to one class of target, but sometimes one still needs to switch to another class of target. Research consortia would be great; they would enable us to exchange our chemical libraries more easily, so we wouldn’t have to reinvent our own chemicals. At H3 we’re building a proprietary chemical library that’s suitable for certain classes of target, including those which have traditionally been considered to be undruggable.
- Have the right assay in place. Many organizations settle too quickly on an established assay. We’re trying to establish a screening platform that’s medium throughput (3,000-5,000 compounds at a time). We can then look at different versions of an assay to find the right one. We need to have the right components in an assay and make sure the assay allows the enzyme to be regulated the way it would be in a cell.
Q: Are you developing companion diagnostics? How do they fit within your drug development strategy?
Warmuth: This is absolutely key to our strategy, which is patient-centric. Every project we’re working on starts with a hypothesis derived from real patients (or, more specifically, patient samples). Then, we always want to explore the effects of a drug candidate in a population related to the population we started from with our original genomic data (we would not study the same patients, but a similar population with same mutation). But we need to identify them with companion diagnostics. This area will get more complicated as patient populations get smaller and smaller. Today, target-centric developers are looking at something that happens in 5 to 10 percent of a certain cancer type, which makes recruiting for clinical trials challenging. As an example, our lead project is on a mutation found in a gene called SF3B1. We know this mutation occurs in multiple different cancer types, but in some it is very rare, in the order of 3 to 5 percent. It will be crucial to have the right companion diagnostic in place to allow selection of those patients once we are in a clinical trial.
Q: Where do you believe H3 will be in five years?
Warmuth: We want to be at the stage where we’re very close to clinical proof of concept. We’ve validated a new discovery paradigm that is focused on human biology, the human cancer genome and new metrics of drug discovery. It would be nice at that time to be able to reflect on the results of the strategy that we have discussed today; to have access to more models, to more closely incorporate patients as part of the discovery process, and to be able to access patients more easily for hypothesis testing.
What do you think of H3’s model? How have advances in genomics affected your product development strategies? Is there a way to digitize this new biological knowledge, or even create tools to help uncover other biological trends? Please share your thoughts with us here.
Like this content? Don’t miss an update.
Leave a Reply