3 Questions With Merck’s VP & Head Of Neuroscience Research
If you looked up Darryle Schoepp’s background on LinkedIn, you could comfortably say that for 30 years he had one focus — neuroscience R&D. But it has actually been longer than that. “Many people probably don’t know that I decided to enter this industry and the field of R&D at age 16 when I came across a Merck R&D pamphlet,” he shares. Schoepp eventually joined Lilly in 1987, the same year the FDA approved Prozac. During that time, he recalls how everyone was investing heavily in neuroscience R&D to develop the next best anti-depressant. But as Prozac, Paxil, and various other anti-depressants went generic, and as companies moved away from developing “me too” drugs, Schoepp witnessed the neuroscience R&D investment pendulum swing the other way. But today, he says, momentum is shifting back to neuroscience. In this Q&A, he gets us up to speed on developments in neuroscience R&D and why there is significant increased interest in developing therapeutics targeting neurodegenerative diseases.
Life Science Leader: What Is Your Thinking Regarding Combination Therapy In Neurodegenerative Disease Drug Development?
Schoepp: Combination therapies are very common in neuroscience. In fact, practicing psychiatrists do combination therapy 75 percent of the time. For example, psychiatrists will add low doses of clozapine, an atypical antipsychotic often used in schizophrenia, to another antipsychotic, because high doses of clozapine have certain risks. So they give low doses of clozapine to get the unique qualities of the drug, but then add a different drug or drugs to get a desired therapeutic effect. This common practice is often done without the support of data from clinical studies.
In the neurodegenerative area, you often have diseases that have multiple pathologies. If you have a patient where amyloid pathology and tau pathology are present, it might make sense to use combinations of therapies targeting both pathologies to hopefully give a better result. This makes a lot of sense scientifically, and there are animal studies supporting this practice. But clinical studies in neurodegenerative diseases like Alzheimer’s can be particularly onerous. You can run underpowered phase 2 safety studies and look for hints of efficacy, but thus far that hasn’t predicted phase 3 results very well. For example, bapineuzumab, a drug being developed by Elan, Pfizer, and J&J for Alzheimer’s, had post hoc data that looked very promising, but it didn’t work. Baxter’s Gammagard had clinical trial phase 2 data and biomarker imaging data, but this did not translate into successful phase 3 data because the phase 2 trials were underpowered. To really see if a drug works in clinical phase 3 studies you have to follow a bunch of people who decline for 18 months to two years, which is expensive and time consuming. The biopharmaceutical industry has spent a lot of money trying to prove efficacy of single agents in treating Alzheimer’s disease, and thus far this hasn’t worked very well. In fact, Alzheimer’s disease drug candidates have one of the highest failure rates of any disease area (i.e., 99.6 percent). What we need are ways to investigate drugs alone and in combinations in smaller studies that hopefully tell us the biology is moving in the right direction, and that there might be some benefit beyond the single agent. But the field isn’t there yet, and we don’t currently have the validated progression biomarkers to better guide clinical investigation. But we may soon have what’s needed to run proof of concept (POC) studies like we do in other diseases. In Schizophrenia for example, you can run a six-week to 12-week study with 100 patients per arm just to see if anything is there. It doesn’t mean that everything will work out down the road, but at least in other indications you can do such a study. As it stands right now, we can’t do short smaller studies and have any predictive power when it comes to delaying/preventing neurodegenerative diseases.
Life Science Leader: Are Genetic Discoveries Opening Up Opportunities For Better Biomarkers In Targeting Neurodegenerative Disease?
Schoepp: Yes. Genetics are starting to tell us there are populations where you might be able to study, perhaps find a benefit in that population first, and then expand. For example, GBA gene mutation populations aren’t that small. And while people with GBA gene mutations have an increased susceptibility to Parkinson’s disease, not every patient with Parkinson’s has that GBA gene mutation. But there are enough people with glucocerebrosidase (GBA) gene mutations that do have Parkinson’s disease, and we know there is some idiopathic link between Parkinson’s disease and GBA gene mutation. So we should be able to get more homogeneous populations of patients with both to run a clinical trial to evaluate efficacy and then expand. Another Parkinson’s disease risk factor is the leucine-rich repeat kinase 2 (LRRK2) gene. Testing this would be harder, but you could try to find LRRK2 cohorts, which is what some researchers are trying test for and if successful, expand beyond that. Science is trying to link these types of genetic factors to everyone with a particular disease.
Neuroscience researchers are also developing new-model constructs for investigating misfolded proteins for diseases like Alzheimer’s, Parkinson’s, ALS, and dementia. Today we are able to put precursors to these misfolded proteins into animals to see the pathology evolve. These are different than the models we had before, and are enabling our ability to run new types of preclinical studies. However, one of the big issues is chemists, who like to optimize molecules for single targets such as receptors and enzymes. In that case, they know how their molecule binds and how it affects a protein or an enzyme. Though we are searching for those types of targets, these newer targets reside in pathways where the best target is not clear. Tau for example, is not a simple target. People don’t have the amyloid equivalent of a beta-secretase enzyme that all pathological forms of tau come from. But we have a pathway and assays, and so researchers are now searching for targets having the most impact on that pathology, like a specific kinase or antibodies that bind extracellular forms of tau and maybe prevent the disease from spreading.
Neuroscience has been evolving into a different science, one driven by modeling human biology. Simply stated we are now using translation of human biology to understand the disease, what are the differences between those that have disease and those that don’t, bringing that information back to develop models, using those models to develop new targets, new ideas, and new therapeutics.
A lot of my previous work focused on, what we thought were predictive preclinical models for psychiatric conditions. Though we had somewhat limited success, one success we did have was in the area of sleep, which seems to translate well across species and you can run a study fairly fast. Suvorexant, sold under the trade name Belsomra, has the right profile to be given to healthy people, and you can pick a dose just from what makes a person sleepy. But insomnia is a much easier problem to solve compared neurodegenerative diseases such as Alzheimer’s. In the neurodegenerative area, we need similar types of translation. In addition, we need to have preclinical models and be thinking about what can be done to demonstrate the human disease biology is there, and how to positively impact things like tau in human patients safely. We believe if we are able to do this in patient populations that have such disease pathology we will have a high probability of being able to help patients. If you give an agent to prevent tau pathology from spreading in an Alzheimer’s patient, using let’s say a tau positron emission tomography (PET) ligand to monitor disease progression in a cohort of people, that might have a lot of predictive power and would justify going into a big phase 3 study. If we have these clinically validated disease progression biomarkers to track the rate at which patients are getting worse, it may also not matter what the mechanism is, because you’re effecting disease progression. Then we’d be better able to investigate combination therapies in a more efficient and cost-effective way.
Life Science Leader: What Is The Link Between Neurodegenerative Disease And Brain Inflammation?
Schoepp: The immune system has pro-inflammatory pathways and anti-inflammatory pathways. There’s a gene called TREM2 where having a mutation of that gene greatly increases one’s risk of Alzheimer’s. This gene is found on microglial cells and thought to be associated with loss of microglial function. There are these receptors that sit on these glial cells responsible for basically cleaning up the brain from cellular debris such as amyloid plaques, typically found in Alzheimer’s patients. Activated glial cells and TREM2 have been associated with amyloidplaques. There is other data (i.e., PET ligands called TSPO which label activated glial cells) suggesting that brain inflammation happens early in Alzheimer’s disease, and might actually be contributing to the worsening of the disease in early stages. Even though it may not be the primary pathophysiology, it may be helping to make things worse. This is somewhat similar to what you see with joint inflammation, where the body stimulates an immune response creating inflammation. It this inflammation becomes chronic it is then a problem. There is a soluble form of this TREM2 receptor that shows up in cerebral spinal fluid (CSF). A recent paper studying a cohort of dominantly inherited Alzheimer’s patients indicates CSF soluble TREM as a biomarker, and they measured it for tracking this inflammatory process. So there’s the new area in neuroinflammation and TREM2 biology to consider as a way of modifying the disease process in Alzheimer’s.