Picture Perfect: Imaging in Drug Discovery and Translational Medicine

Imaging is taking on new roles in clinical trials, discovery, and even patient stratification.  

By Allison Proffitt 

February 7, 2012 | Imaging in drug discovery has been rising to the forefront of conversations more and more recently. The IT & Informatics Best Practices winner for 2010 was Novartis’ Image Analysis Interface and ImageECD, while the Oyster Imaging Collaborative Portal from Pfizer won an honorable mention in 2011. Managing editor Allison Proffitt spoke with Ken Kilgore, Director of Immunology Pharmacology at Janssen Research & Development (formerly Centocor Research & Development), a Johnson & Johnson company, about how—and why—imaging’s role is changing in drug discovery.  

Bio•IT World: We are hearing a lot more about imaging in discovery. What’s changing?
Ken Kilgore: Yeah, that really fits in with what I’ve experienced. There was a significant amount attention 10-12 years ago in the industry. The industry put in a tremendous amount of money into imaging, purchasing MRI units, Computed Tomography (CT) scanners or something similar. But despite all this funding and dedicated space and resources, staff, etc., very little materialized except a lot of pretty pictures that people didn’t know exactly what to do with… so it went quiet.

In the past two years, we have seen another explosion, not so much in the machinery itself, though that has obviously made immense advances, but in the creation of a number of smaller companies that are focused on imaging and helping out with the technologies around it. Also, the post-processing software has caught up. No longer do we have to hire a radiologist or spend a day scanning one image and looking for changes… From this [current] software we’re getting meaningful, quantitative measurements we can now graph, which is really important because that’s what people understand. When you put that picture up, you can put a graph next to it and they’re going to be able to read that graph and then relate it to the picture. This post-processing software is what allows us to conduct more meaningful research… And we can also now “train” the software, that is teach it what to look for and it will do it. And that’s something that never would have happened ten years ago. It just wasn’t there.   

CT image of mouse normal (left, top) and fibrotic (left, bottom) lungs. Quantitative analysis (right) allows for the determination of the degree of fibrosis   in three normal lungs and three fibrotic lungs.

Here at Janssen R&D [my group is] working in the field of immunology, which includes autoimmune diseases like rheumatoid arthritis (RA), also pulmonary diseases like asthma, chronic obstructive pulmonary disease (COPD) and pulmonary fibrosis. The use of imaging initially focused on arthritic disease. We were actually imaging for conditions like RA, osteoarthritis—that’s been around a long time—even in preclinical drug discovery. Where I think Janssen R&D is really taking the lead is in pulmonary indications, and that includes COPD and fibrosis, where we can model those diseases and look at them in a mouse, and get high-resolution images, and, importantly, quantitative data from that. What’s important is that we’re able to see through these images if a drug is working in this particular model [or not]. That capability is really a valuable tool because it allows us to take that data in the animal models and then… translate that data up into the human. Because of what we can do with imaging, now we can start making those connections: Is what we’re seeing in these models what we’re seeing in the human? If we’re seeing it in the animal models maybe we should look for it in the human… What we’re seeing in the preclinical arena is really what we’re seeing in the human, and that strengthens our preclinical drug discovery efforts.  

Are you using imaging more to study disease progression or drug efficacy?Both! In the medical community, we have been trying to treat diseases that we don’t fully understand. That is, we don’t know the exact pathology of these diseases we’re trying to treat. But what we’ve really taken a concerted effort to do—and imaging is a part of this effort of course—is to start to understand the disease progression. We can do this at the molecular level, at the cellular level, and now through imaging we can do it in the whole organ, the whole body. This helps build our understanding of the disease, and once we know what those changes are, it really helps strengthen our efficacy. [In animal models], we know what’s correlated with disease and what’s just [aging or] growing up.  

Can you give me an example—a picture, if you will—of how this works?Sure, let’s look at it from an efficacy standpoint in a disease model of fibrosis. Our group characterized the animal model with these imaging platforms; we know what to look for. In classic models of fibrosis, what you’d have to do is sacrifice a group of animals and look at this histologically. With the advent of imaging and subsequent improvements, we can take a set of animals and do many more treatment groups, do a dose response, we can image the same set of animals over the course of the disease. So if the model takes six weeks, we can analyze those same animals over the entire six weeks and track real quantitative changes and follow each individual animal over time as opposed to sets of animals. It really makes the data stronger.

We’re also able to conduct research with fewer resources and much less drug because we’re studying fewer animals. Then we can use clinically relevant endpoints. The same endpoint they’d use in human fibrosis, we can use in the animal model, which is much more translatable. It allows us to follow longitudinally what’s going on. What’s really important there, and we can do some of this biochemically, is the translatability factor. That’s very important. Our end goal is really to help our clinicians design clinical trials.

You mentioned there are many smaller companies now pushing imaging forward. How does Janssen view and work with those groups?Let’s say, for example, we’re looking at a lung sample. We collaborate with another company whereby we ship the sample over to their researchers and they’ll ‘dip it in their secret sauce’, so they’re literally immersing the lung inside a fluid, and then they send it back to us. What they call the secret sauce really enhances the sensitivity, allowing the imaging machine to detect changes much more readily. The pictures are much more clearly defined.

We’re still in a growth phase, and there are a lot of collaborations being set up. We’re not just paying small companies to do things; we work together and through our collaborative efforts with these companies we try to enhance their technologies to ultimately benefit patients. 

With the growth you’ve seen in the past few years, what is the future for imaging in drug discovery?We see imaging as a very important technology. It’s now something that’s readily doable and just as importantly, something that’s clinically relevant and translatable to the human. Our team is very interested in translational medicine and the promise it holds for patients. We see that imaging is technology that’s translatable to humans, but we’re also looking down the path to see what additional efforts imaging may be able to support. For example, maybe in patient stratification, where we are able to differentiate diverse patient populations based on clinical imaging and then complete clinical trials within those stratified patient populations. There has been tremendous progress in the software that’s being used to quantitate imaging results in drug discovery. I believe there’s much more room to grow there and that we’ll be able to get to even greater degrees of sensitivity, both clinically and pre-clinically.

When I look at where we’re going to apply this, it’s in the future. Maybe we will complete imaging earlier in the course of a disease and follow it, just like we did in the preclinical stage. We’re gaining understanding of the human disease now and how the changes are occurring, and then maybe that ultimately will allow us to reach the ability to stratify patient populations, which has an enormous impact on our ability to run smaller clinical trials of shorter durations. Seeing those efforts come to fruition, I believe the investment in the technology will have paid off.