Curran: That’s a great question. I think personalized medicine, when many people use the phrase, they’re meaning to indicate a genetic test, DNA-based. Where suitable that’s a wonderful scientific opportunity because DNA, except in the case of cancer mutations, is a very stable marker. If you create a robust assay for a DNA marker it should be a good clinical test. So there are advantages to looking at DNA from strictly a technology and a handling perspective. When I use personalized medicine as a strict term—not colloquially—I mean to indicate a genetic test, or a DNA based test. And that would truly be personal. That would be an individual patient’s genome dictating what their likely drug response or safety profile was meant to be.
I like to use the term stratified medicine to indicate that we might be looking at different biosamples and biomatrices besides DNA. For instance a proteomic signature, or a gene expression RNA signature, or perhaps a metabolomic signature. Any of the measurements that are available—if you can find the appropriate biomolocule to measure—could be a good biomarker, and if the technologies are robust enough that could be translated into the clinic.
If you look at the history of this space, personalized medicine in many ways is synonymous with pharmacogenetics, which can be used to indicate DNA. We’re using “stratified medicine” in some ways to indicate that we’re agnostic to the assay technology. We want to find the right marker with the appropriate power and the appropriate robustness that can translate the scientific finding in a meaningful way to enable clinical decision making.
So in some ways, Allison, I think it could be semantics, but I think there’s enough history in this space, and as you said, enough buzz, that maybe the terminology could be a little more precisely used.
In which areas is a stratified approach most appropriate?
We’re particularly excited about several areas. We have a rich history in what we consider the core inflammatory diseases of rheumatoid arthritis, psoriasis, inflammatory bowel disease and some of the related syndromes, and we’ve been very successful in introducing novel therapeutic solutions in those diseases. And we think that in many ways the next evolution will be better directing those therapies...
And if we look at the field of respiratory and pulmonary disease, I think we’d be hard pressed to find anyone working in that field that thinks asthma is one disease, or COPD is one disease. And if we look at the therapeutics in use today, they’re generally effective in just a subset of individuals. So from a stratified medicine perspective, that strongly suggests that to have a novel efficacious therapy, you also need to couple those therapies to a diagnostic that lets you select the patients that are most likely to benefit. And as you know, that’s the whole story of personalized or stratified medicine. I think the difference between respiratory diseases and some of the other conditions we’ve worked on is that biomarkers will almost be, by default, a necessity in the respiratory space.
You chose to focus this consortium on rheumatoid arthritis. Why is RA a good fit for a stratified approach?
In rheumatoid arthritis there are a number of different treatments, different biologics: there are anti-TNFs, there’s an anti-B cell, there’s a co-simulation modulator, there’s anti-IL6, there’s anti-IL1. All these are biologics and they’re all suitable treatments. The anti-TNFs tend to be used first line, and there are five of them now. When patients respond, they tend to do very well. They’re quite nice drugs. The patients respond well and they have a high quality of life, when they don’t respond—and some patients don’t respond—then you have a situation where you would prefer to use another drug… This is an area where there are therapeutic options, good therapeutic options, and it’s an area where you can see that personalized medicine or stratified medicine could be very impactful.
How did the BATTER-UP consortium get started?
Almost five years ago now, several of us in the field put our heads together and asked: Why aren’t we making progress to the end game to validate predictive signatures [for rheumatoid arthritis] and move them into commercial diagnostics and really help patients pick the right drug? Why do we have reports in the literature that go un-replicated or un-validated?
And in thinking about it, the reason became very obvious: it’s actually difficult to collect the appropriate patient material with good phenotypes to validate these markers. So you have a number of really good initial reports that aren’t followed up on because the individual labs that made those findings don’t have the resources, or have access to a larger patient cohort, in which they can validate the marker.
So we thought, this is a perfect opportunity for several of the companies that are interested in this space and dedicated to better care for the patients, to potentially come together and facilitate the collection of a large cohort of individuals starting an anti-TNF drug.
We can then follow those individuals for a reasonable period of time, see how they do, define response, and then ask the questions: do any of the biomarkers that have been reported in the literature replicate, do they validate, are they real? Can these markers be [converted] to a diagnostic?...
It’s almost the perfect storm, where the technologies have reached the point where they’re yielding results that need to be validated and there was a clear gap, which was lack of a large population or large cohort to replicate findings and to drive new ones.
The consortium consists of Biogen Idec, Bristol-Myers Squibb, Centocor R&D, Crescendo Bioscience, Genentech, Medco Health Solutions, Regeneron Pharmaceuticals and Sanofi-Aventis. How did the group come together?
I have to admit that I’m amazed that we were able to put this together because it’s fairly unique to see large companies collaborate, but it’s a wonderful story where companies said, Yep, let’s get together and let’s pool some funds and let’s make this happen.
Medco is a formal part of the consortium as a way to access patients that are initiating treatment. What Medco does is they provide essentially a conduit to physicians. We wanted to do this study in the community. We wanted to reach out to community practices, we didn’t want to run a clinical study or a clinical trial like we normally do, where we go to a few sites, and get a well-characterized academic population.
We wanted to go out into the real world and make sure that any biomarkers that emerged would be applicable to the population at large. So we felt it was important to have a broad reach...
Our key opinion leaders are Michael Wineblatt from Brigham and Women’s and Peter Gregersen from New York Long Island Jewish Hospital… and we have a scientific advisory board as well.
The consortium is funded by the partners and the partners are taking on the responsibility for organizing it, but we don’t intend this to be anything other than an opportunity to validate predictive biomarkers in RA and see if we can make this a reality for patients.
There is a tremendous amount of organization to make a consortium work. Personally, I needed to be convinced that our partners and my colleagues and peers were committed to a successful outcome—and we’ve assembled a group that is. From that point forward, it’s very easy to get the enthusiasm for it... Let’s make it happen. Really no magic. It’s just a matter of saying let’s do it, and get it done.
Where does BATTER-UP stand now?
We’re in the process of collecting 1000 patients. Enrollment changes daily, but we’re approaching triple digits now. It’s very early. We just started the protocol middle of last year and studies take time to ramp up…
In terms of design, we’re taking our initial phenotype in the 12-14 week period. If we’re successful at predicting response at 12-14 weeks, that would be dramatic. Because remember, today, you can’t predict response for anything for any duration of time. 12-14 weeks is the time when physicians that are treating individuals with RA in their practice are looking for response. If they’re not seeing response at that time, they’re generally thinking about switching drugs.
Having said that, the intellectual discussion then becomes, well 12-14 weeks is not that far from initiation of treatment. You’re talking about three months. Wouldn’t you want to predict response at a year or two or three? And yes, we would! However, step wise—if we can do 12 weeks, then we’ll have confidence to tackle six months and a year.
BATTER-UP is a very good and very well-defined experiment with grounds in both regulatory endpoints and clinical practice. So we believe we have set the stage for success.
Will there be a BATTER-UP 2 if this first phase is successful?
I don’t know. We talk about it now. This is a very deliverable-driven group, and we want to successfully complete BATTER-UP 1. But if we’re successful and everybody involved finds value in it, I could easily see BATTER-UP 2 where we look not only at the anti-TNFs but at other drugs that work in RA. So there’s definitely potential for this to be broadened to a larger patient cohort or perhaps longer time points: maybe 6-month response or maybe 12-month response. But the focus of the consortium is solely on executing BATTER-UP 1 and making sure that we do that successfully.
We certainly don’t want to claim success for this model and for BATTER-UP until we can demonstrate it. But the group of companies coming together to address the problem in a precompetitive manner, it’s not unique, but it’s rare. And I think if we are successful, you’ll see other companies doing this sort of thing in different diseases.
This article also appeared in the March-April 2011 issue of Bio-IT World Magazine. Subscriptions are free for qualifying individuals. Apply today.