Professor of bioengineering, chemistry, and medicinal chemistry
University of Illinois
Tony Hopfinger is a professor of bioengineering, chemistry, and medicinal chemistry at the University of Illinois at Chicago’s medicinal chemistry and pharmacognosy department. He is also co-organizer of the ADMET 2 conference, which will be held next week in San Diego.
With at least two companies launching ADME/tox products this week at Lab Automation in San Jose, Calif., (see page 1) BioCommerce Week spoke with Hopfinger about the state of the ADME/tox market as tools companies respond to pharma’s need to fail compounds early, more cheaply.
How do you describe the state of ADME/tox today?
Basically, for the last 10 years, myself and a lot of other researchers have been trying to find ways to make accurate predictions on ADME/tox properties. In the preclinical phase of drug development, the basic focus was always potency on the therapeutic end of the drug. Only after you got your compound that had the preclinical endpoint values for therapeutic potency, did people pay attention. What was always clear, but nobody admitted it, was that you spent two years, and the damn thing was not soluble, or was toxic, or not orally active, and you ended up throwing the thing away. That became cognizant about 10 years ago with [pharma] management due to the spiraling cost of developing a drug.
10 years later, what is the state of ADME/tox?
The ADME tools today, experimental and computational, aren’t that good, and you don’t see a lot of successful vendors on the computational or experimental side. It is very difficult to develop an equation. Everything has to be tailored for the particular class of chemistry you are looking at. This has never been quite appreciated as fully as it is for the ADME properties. There are a lot of the tools out there, in-house developed by pharma, or commercial vendors, tailored for the particular chemistries being dealt with in pharma. To develop tailored screens, computational or experimental, is very expensive. To get things that will work for solubility, blood-brain barrier, metabolism, pharmacokinetics, you need dozens of screens, and to have valid ones for your chemistries is an enormous task.
Where is innovation coming from in ADME/tox?
In the field of ADMET, that is a wonderful question. There are four players — the government, big pharma, the vendors trying to develop things, and of course academia. At this point, I don’t think any one of those four is dominant. In terms of funding and novelty, they are all about equally bad. The only thing that may be going on with the government, is that they have recognized this is important. There are some [requests for proposals], grants, set aside to do research and development in the ADMET area.
The thing that gripes me is that the NIH has had this philosophy that ADMET-type research, practical drug-development-type research, was something that they didn’t want to deal with. That seems to be changing. Maybe it’s the realization of how expensive it is to produce new pharmaceuticals, and that maybe it can’t be done by big pharma itself. Living here in Chicago, we have Boeing here and they compete with Airbus for big jet contracts. But, Airbus gets government subsidies, at all levels, including R&D from universities. You can make the case that American pharmaceuticals are fighting western Europe and Japan, and in those countries there is more aid to the pharmaceutical industry than here.
I’ve had a software company for 25 years, a dual-hat situation, where the work is done at the university, and at other universities, and by consultants. We do a lot of applications work, which is much more profitable right now than new software. Academics can do development cheaper and get support from the government. Where we fall down is in making it reliable and simple to use. What should be going on is a marriage between the vendors to create the front end, and the academics to create the software you should see.
You are organizing the second ADMET Conference in San Diego next week. What is your goal there?
We want to open the dialog between the big four organizations and we have tried to keep it a research-type environment from the presentations. A lot of the other [ADMET] meetings are a forum for funding with smaller companies making a pitch to an audience of venture capitalists and big pharma. We are trying to religiously maintain the tech sessions as science and make it cheap to get students and the academic community involved. For my point of view, one of the reasons I wanted to be part of it is that the academic community was being locked out of this type of area, other than the American Chemical Society.
What kind of attendance are you expecting?
We aren’t getting 200, and it’s in San Diego in February. But, other meetings aren’t getting them also. Maybe there is an economic reason for it, as people haven’t seen huge breakthroughs in ADME. There is a lot of ADME work being done, but it is using known and established techniques. There is not a great deal of new stuff there.
What would be your ADME/tox wish list?
Given a start from a baseline of zero, there is a lot of stuff that is going to be helpful. A lot of new equipment; predicting lipophilicity of some type of membrane transport, or maybe some type of products that deal with metabolism by seeing how a compound might break down. All of these things are good, but the hard part is that you have this general tool that can make these general measurements.
More often than not, those don’t correlate to your specific problem — they have to be tailored for your problem. A lot of people are buying stuff, from Bio-Rad and other people. But I don’t come away with the feeling that they are doing very well. Maybe people are having financial reservations. At least from last year’s meeting, I had the perception that these things might not be overly useful. Maybe because there is an increasing number of options, people are kicking the tools and there is still a lot of that going on.
Metabolomics seems to be on an up-ramp. Is that hopeful for ADME/tox applications?
Metabolism is more of a black box, from a computational point of view. I’m not saying it is not an important way to go, developing databases for mass specs, and keeping track of the metabolic products and sampling a drug at various points in time after being given to a biological system.
But, it presupposes that you have made a compound. People are making progress, and there are a number of software packages out there that can do predictive metabolic type studies.
Are informatics helpful?
There I become emotional. To do computational simulation of biomolecular systems, that is an enormously difficult undertaking, and you need to know about quantum mechanics, and you go on and on. When the field started, people worked about the huge details of how a ligand might bind to a receptor, but over the years, the difficulties of that type of modeling became more apparent, and didn’t make a huge amount of progress.