At A Glance
Name: Steven Burrill
Title: CEO of Burrill & Company —1994 to present
Background: Technology and sciences consultant, Ernst & Young — 1966-1994
Steven Burrill, CEO and founder of the merchant bank Burrill & Company, has watched the ups and downs of the biotechnology industry for the past 35 years. A recent bank release, “Burrill’s Biotech Outlook 2005,” looks back on 2004 to assess last year’s predictions, and forward to an optimistic 2005 for pharmacogenomics.
Burrill sits on the boards of directors of several biotech companies and non-profit organizations, and on the advisory boards of four universities.
In Burrill’s Outlook 2005, you predicted good things for pharmcogenomics. What do you expect to happen to pharmacogenomics this year, and why?
I think we’re going to move very dramatically from the one-size-fits-all drug world that has yielded blockbuster drugs, and the goal of big pharma and others to create new blockbuster drugs, fairly rapidly to a world of more personalized medicine. And through personalized medicine to both predictive and then preventative medicine.
And as we make that transition, the biggest drivers for that transition will be two fold. One, the enabling technologies, i.e. pharmacogenomics and other enabling technologies that will [allow] us to identify particular patient populations who may be responders or repositories for certain behavior — existence of certain mutations or whatever.
And other technologies or informatics that enable us to discern what might otherwise be an adverse drug reaction or side effect or a non-responder in some way.
So, as we begin to make this transition — initially in theranostics, ultimately in preventative care — we will see dramatic value creation and use of, and broad-based work with, pharmacogenomics.
That transformation will be driven also, not just by the enabling technology, but the other big driver will be the payor community. Because, if you say, “What’s really going to drive change?” It’s the ability of those who pay for medical care to have a way to know that the medical care they’re paying for is working, or to discriminate out those patients for whom it won’t work, therefore they can reduce costs.
The two big drivers here are going to be the payors at one end, and , if you will, new technology, of which pharmacogenomics is a critical piece, at the other.
The force driving payors to push pharmacogenomics — is that a cost-benefit calculation?
There are two or three fundamental things.
One, roughly 50 percent — as the New England Journal of Medicine finds the numbers — of the medicines prescribed today don’t work for the patients for whom they are prescribed.
Some of that has to with compliance, some of that has to do with idiosyncrasies of “you versus me,” a lot of it has to do with the “one size fits all.”
Secondly, the fourth-largest cause of preventable death in this country is adverse drug reactions.
And thirdly, if you look at the cost of healthcare today, 80 percent of our $1.4 trillion of healthcare costs, is for chronic care. So, if you are a payor, and you want to reduce your healthcare costs, you can do it by reducing your large cost for chronic care, which leads you to using these technologies to identify these things earlier, to eliminate the high cost and the difficult circumstances of adverse drug reactions.
So if you took any $10 billion drug, and you just took that 50 percent and said you could eliminate $5 billion of that cost, and Mark McClellan at [the Centers for Medicare and Medicaid Services] was paying that bill, you can be sure he would be an applier of that technology.
So, it is cost-efficiency, it is quality healthcare.
It also, interestingly enough in my view, is the vehicle by which we will bridge this gap in the FDA/pharmaceutical world, where the FDA is getting beat up because of Vioxx and Celebrex and others, without being able to identify populations of patients who could have been at risk, and the use of this technology that may very well be able to identify those patient subpopulations. And therefore, in the drug-approval mechanism, we can label that away or deal with it at the approval level.
And perhaps even create theranostics, where you get your Rx approved as long as you’ve got a tandem Dx that can identify the responders.
So all those changes are going to come out of — broadly speaking — rapid change during 2005 and 2006.
The recent P450 chip with Affy and Roche, the company XDX [which has a] new diagnostic [...], Genomic Health — there are a lot of diagnostics built around these these new technologies that are beginning to make their way into the marketplace. And while priced more like a pharmaceutical than historically like the diagnostics have been priced, [they] are enormously more valuable in the context of cost efficiency.
Do you see any signs that payors are seeing pharmacogenomics and its related diagnostics as cost-efficient things to do?
Absolutely. And they have the economic incentive to apply them wherever they can, if they believe that it can either improve quick care or reduce costs.
What are some of those signs?
If you go to Kaiser, you can identify cancer earlier than we do today. So, you’re now waiting for a patient to present themselves with stage-IV problems, but you know you’ve got some abnormal cell growth early, and that tells you you have a pre-cancerous and early cancer condition, that if we did certain things, we can deal with it .
The cost effectiveness of this could be extraordinary.
So the payor, somebody like CMS at a government level, or somebody like Kaiser at an industrial level, has every incentive in the world to spend money for things that are going to reduce the overall cost.
The place it doesn’t work is if the cost-benefit is outside the patient population, or outside the time you’re going to have that patient. So, if you’re going to only have patients — on average — five years in your program, and the benefits are in year 15, that’s probably not going to be done.
But if you take a Kaiser plan, where most people by and large are in forever, or at least for the long term, or you take an employer-driven plan, where they’re assuming these people are with them, there’s a lot of incentive on the part of those payors to apply all of this technology.
And heretofore, we didn’t really have a technology that was applicable.
Do expensive tests face different pressures in the face of expensive drugs?
I do think that the mix of costs will be important, my belief is that the relative value of diagnostics will go up dramatically, and the value of therapeutics come down. If you stipulate for a minute that you live in a world where therapeutics are high-value, high-margin products and diagnostics are low-value, commodity-based products, I think that relative value would change fairly dramatically.
If it’s a diagnostic that’s going to identify the patient responders, or a diagnostic that’s going to tell you whether there’s compliance, and it’s working, and it’s a diagnostic that the payor is going to be able to use to make sure that you’re a responder and it’s going to be an effective treatment, you’ve got to believe that these new molecular diagnostics are going to come into the world with a very different value proposition than they have historically. So, I see examples of that coming into the market place today, where you see diagnostics — are they going to end up with therapeutic-like margins?
You just launched a new $300 million–$500 million biotechnology fund called Life Science Captial Fund III. How much of this will go to Pharmacogenomics-related businesses?
Well, the focus of this fund is in the sickness-to-wellness paradigm, and I’m kind of a passionate bigot about this move to personalized medicine, and then to preventative and predictive care. So, my guess is that we will apply a large amount of that in that space.
So, this isn’t going to be a fund that funds all the late clinical development compounds in the hope that there’s some value building as we go from Phase 2-b into 3, or something like that — it will be more broadly applied across that changing world of medicine.
We’re going to invest roughly $15 million–$20 million per company, with $7 million–$10 million in the first round, and you put an equal amount in your back pocket and support the company later on.
We’re probably going to do very little in the services area; we historically haven’t done much in the pure-play informatics area — but that gets a little fuzzy as you get to pharmacogenomics — but we are going to be investors in technologies that are important enablers of this change.