Health knowledge made personal
Join this community!
› Share page:
Go
Search posts:

Personalized medicine for the brain. A discussion with Brain Resource’s Evian Gordon (transcript)

Posted Feb 13 2012 10:40pm

This is the transcript of my recent podcast interview with Brain Resource Company chairman Evian Gordon.

Williams:            This is David Williams, co-founder of MedPharma Partners and author of the Health Business Blog.  I’m speaking today with Dr. Evian Gordon, executive chairman of the Brain Resource Company .  Evian, thanks for being with me today.

Gordon:            It’s a pleasure.

Williams:            We’re going to talk about personalized medicine for the brain. So first off, how is personalized medicine for the brain different than other kinds of personalized medicine?

Gordon:            Well it’s no different at all.  The goal essentially is to find biological markers that can accurately predict treatment response.  The difference is one of more pragmatic reality. Most of the findings in personalized medicine that are well learned have come out of the area of cancers and HIV/AIDS.

The most common examples cited are Herceptin for breast cancer, Selzentry for HIV/AIDS, Gleevec for leukemia, Iressa for lung cancer and Erbitux in colon cancer.

Even though there are a small number of findings, there are a growing number and it is rather surprising that they have been confined to the cancer area in the main.  So that’s been the biggest difference, but that I think is one of simply where the focus has been and where the investments have gone into. It’s cancer.

Williams:            Now speaking of investments, I know that you are involved in the iSPOT study , which I understand is a large study that is related to the brain and personalized medicine.  Can you tell us a little bit about that?

Gordon:       Sure.  This is a study from a European biotech. It’s a $20 million study. The goal is essentially to look at psychiatric disorders starting with depression and ADHD. The principle is to look beyond just the molecular findings –all the findings so far in cancer and HIV have been molecular.  The current word, as you probably know, is “panomics,” meaning everything from genomics to gene expression to metabolomics; everything that moves at the molecular scale.

And while that’s absolutely noteworthy and important, in the brain where most genes seem to be involved or 80% of our genes possibly involved in psychiatric illnesses, it seems unlikely that genes alone or any form of panomics are going to be sufficient to sensitively and specifically predict treatment response.  So what we’ve done is set up the first global standard to measure both molecular, but also everything else about the brain; the brain structure, functional MRI, electrical brain function, cognition and real world outcomes in addition to standardized clinical workups to see if by combining genes and brain markers we have a better chance of revealing some of these underlying biological disturbances that can predict treatment response.

The little catch is that it requires significant numbers.  By significant numbers I mean thousands.  This study is studying 2,000 patients.  We’re just looking at the first 1,000 at the moment and in the process also of very efficiently bringing integrated analysis facilities to really mine for the best biomarkers that predict treatment response.

It’s a fascinating phase and the principles are no different to any other aspects of biology. We have standardization.  There are 20 sites, ten in the United States and ten in Europe and Australia where the patients have been drawn from.  We have the power of standardization of all measures; hardware, software, ways of analyzing, but also the integration of all these methodologies and then the power of numbers.

Williams:            Why is it so hard to predict treatment response with psychiatric conditions?

Gordon:            Well I’m not sure that it actually.  It’s just that the current model is pretty much the opposite of looking at really standardizing the diagnosis of using signs and symptoms.  It was a wonderful effort when it occurred in DSM.  The treating was a shift forward from psychotherapy type analysis and trying to find a standardized way for diagnosis.

It turned out that signs and symptoms, if you take depression for example, asking questions like, “Did you sleep poorly?”  “Have you lost your appetite?”  “Have you lost your ability to experience pleasure?”  These have not turned out to be the sorts of subjective questions that have done well in predicting who will respond to which antidepressant and that’s why the results, the data from anti-depressants have been so poor.

It would seem in the whole of biology and the whole of medicine that we are essentially redefining medicine based on biology and essentially psychiatry is now entering that phase, that paradigm shift of seeking to find the biological underpinnings, which can hopefully be more accurate and objective in predicting treatment response than have been signs and symptoms.  While they have a value in diagnosis, they have shown to have a much less value in predicting treatment response especially at the individual level.

Williams:            You mentioned standardization as an important component of this iSPOT study and the overall approach.  Talk a little bit more about why standardization is important here.

Gordon:            That’s a good question David.  We’re an international consortium of medical scientists who are looking to study the whole brain as a system and find these biomarkers.  I suppose the context is that very little happens without standardization.  Very little of scale happens without standardization and that’s not just about biology.  If you look at all the major projects across history that have shown really big insights, they’ve required scale and they’ve required standardization.

One can only presume that when you’re dealing with a complex system, that’s just so that you can really compare apples with apples and not continually having small numbers of subjects with small effect sizes and the huge number of confounds so that you simply can’t compare one study with another very easily.

Standardization gets rid of that problem and allows you to benefit from the power of numbers. And if there is an effect size, you really know that it’s real and reproducible and you’re not distracted all the time by thinking that it could be because a paradigm was a little bit different or the analysis was slightly different.

So if you move the confound in that regard, it’s not the panacea of all aspects of finding biomarkers and certainly not the panacea for finding mechanism.  Sometimes these things are found serendipitously by having a great diversity. But if you can standardize on a global scale and get the power of numbers, it’s one way in which major inroads have been made in other areas where this has been attempted and that’s the reason why broad databases are of course coming back into fashion in science and systems are coming back into fashion in science which had become so fragmented and siloed and specialized.  It’s really an attempt to bring the whole back into tying up all the wonderfully important thousands of details and specializations.  Standardization is the glue that essentially does that.

Williams:            Now this iSPOT study as you mentioned is quite expensive; $20 million, a large-scale effort funded by a biotech company.  I’m sure it will produce a lot of interesting findings, but when you come right down to it, do you believe that there will be a business case for the use of some of the results of the study in pharmaceutical development or elsewhere?

Gordon:            I think that given that there are no current claims with the FDA on the brain and so many, relatively speaking on cancer, I think that it could potentially open the floodgates. We’re using three drugs by the way that constitute about 40% of the anti-depressants used in the United States.  A $6 billion per annum spend and if we can find any biomarker that either predicts if you respond at all or if you respond to one of them preferentially, can you predict that?  Or can you predict side effects and who shouldn’t go on that drug?  Or can you predict who gets better but then recurs?  There are a lot of predictions that can be made that are very valuable and have massive clinical validity in a sense if they work.

Once that proof of concept has been derived and registered and the claim lodged with the FDA where we can publish and replicate it, that would potentially open the flood gates to biomarkers being exploited more widely as has been the case after Herceptin with cancer.  So I think it’s a pretty pivotal time to see whether Brain Resource or iSPOT or somebody else can achieve the first landmark and biomarker. Follow-on effects are considerable and they expand the current model of DSM very dramatically.  DSM themselves are trying to incorporate biological markers, but if you look at the draft of DSM-V it doesn’t have a lot of them.

Certainly NIMH have done something very bold in my view.  They have put out a document called RDoC , which is the beginnings now of having domains that are not DSM based that can start moving towards a biological frame of reference. FDA of course is shifting very dramatically toward personalized medicine.  So that confluence of activity I think is a pointer it’s just a matter of time and solidness of the biomarkers that are found.

Williams:            Explain a little bit about the Brain Resource Company itself.  What do you do?  Who are your customers?  What are you trying to achieve?

Gordon:            In terms of biomarkers, by having this platform, this standardized platform and having attracted this study and numerous others, what we’re trying to achieve is setting up one of the landmark ways of finding these biomarkers, companion diagnostics with the key drugs used in psychiatry and then partnering with either pharmaceutical companies or licensing out the biomarkers to payers where clearly there are huge cost savings for people to actually get drugs who are going to benefit from them.  So essentially that model, the monetizing model of the business would be either through pharma or through payers. We’re already in discussions with both in that regard.

The third possibility in this is to use this platform for drug discovery.  So instead of looking at drug discovery from a molecular level only in the microscopic scale to additionally look at the whole brain as a system and look at the kind of circuits that seem to drive the brain; circuits associated with fear and safety like the amygdala fight or flight system that is so critical to the way the brain processes everything in terms of first and foremost minimizing danger.  So looking at circuits like that in the amygdala, medial pre-frontal cortex or how specific chemicals like serotonin work.  So really teasing out those circuits within the brain associated with specific neurochemistry and explicit molecular variance and using that broader insight to also help develop and discover new drugs.  That would be the third tier of the way we see Brain Resource operate.

We also draw from the same insights about the brain to empower people with insights about their brain and how to train themselves on the web in the aggregation product called mybrainsolutions.com. The principle is the same.  It’s about aggregating information to either –on the medical side, find biomarkers that can improve treatment and on the self-empowerment side for people to use those brain insights to train their own brains to be more effective.

Williams:            Well it sounds really fascinating and that a lot of progress is being made already. But also it sounds like we’re really still in the infancy of brain science.  Ten or 20 years from now where will we be in terms of insight about the brain and how it will be used?

Gordon:            Our hope is that in the next 12 months we’ll be peppering the FDA with claims about biomarkers from iSPOT both for depression and for ADHD. We have a number of reasons to feel confident that we will have the biomarkers even from the 1,000 patients and 150 ADD patients.

Things are done incredibly slowly in medicine but this has been a 30-year germination phase from inception to now for us in setting up this standardized platform and database.

I’d say in the next year we would expect some tipping point in terms of biomarkers, potentially.  Realistically though things always take three times longer than one expects as a rule of thumb.  Hopefully within the next three years there will be some brain based biomarkers that will have been replicated and be accepted as clinically meaningful.

After that point I think that there’s going to be an inexorable –probably slow– shift of how to reconcile this biological tipping point of how the brain and biology is impacting psychiatry with the current signs and symptoms and more of a pragmatic clinical diagnostic focus and consumers getting empowered and reconciling them into this equation.

I think it’s going to take five to 10 years to find the right mix. And certainly when it’s all about who pays for what being such a critical variable, that’s going to be another factor as are many of the privacy and other issues.  I’ll say a year to three years for a tipping point and three years to 10 years to really shift the balance of our understanding and our clinical usefulness of this biological information about the brain.

Williams:            I’ve been speaking today with Evian Gordon.  He’s executive chairman of The Brain Resource Company.  Evian, thanks so much for your time.

 


Post a comment
Write a comment:

Related Searches