About this Author
College chemistry, 1983
The 2002 Model
After 10 years of blogging. . .
Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases.
To contact Derek email him directly: email@example.com
January 29, 2004
Alex Tabarrok over at Marginal Revolution is proposing an interesting idea: He's suggesting that the lifetime of a patent be adjusted for the R&D costs that went into the ideas behind it. Things that took little effort would have a short term until expiration, while those who went through expensive slogs would have a better chance to recoup costs.
That's an idea that had never occurred to me, by which you can tell that I'm not an economist. And I can see the appeal - but I can also see the potential for abuse. For a while in graduate school, I shared an apartment with a fairly doctrinaire libertarian. We had quite a few discussions, as you can imagine, but I found that they often fell into the same low-energy state. He would point out the benefits of some scheme, and I would agree in the abstract. But I'd then imagine how it would have worked out in the northeast Arkansas county I'd grown up in. Theory and practice tended, in my view, to diverge.
As I think they would here, too. What I think this would do is create a powerful incentive to hocus one's R&D figures. In the drug industry, each extra year of patent protection can well be worth hundreds of millions of dollars in revenue, which is an incentive indeed (with hot fudge and extra sprinkles on it.) It's not like our drugs come cheaply to us now, but I can see that for an anticipated blockbuster, ways would be found to ensure that the patent racked up the longest possible life. You already see that happening at the back end of patent terms, as the constant battles with generic companies illustrate. Every possible reason is trotted out for why a given patent is invalid, and every possible means to preserve or extend the patent's term is invoked in response.
Frankly, we have enough trouble working out how much things cost in this industry already. I mean, we can see how much money we're spending, but assigning specific amounts to specific projects is harder than it looks. The number of people working on a project fluctuates constantly, for example, and many of them have other simultaneous responsibilities. Many supplies are bought in bulk and shared between projects, and all the analytical work is done on large capital-budget machines that are used constantly for everything. I've seen all sort of schemes to track costs by project, but all of them have a voodoo component.
Another problem is that the big money-making drug is only one part of a large patent, which would typically exemplify dozens or hundreds of separate compounds. Naturally enough, some of these took a lot more effort to discover, make, and test than others. If keeping track of costs per project is tough, costing things out by compound would be insane. Then there's the problem of multiple patents protecting the same drug. You'll see a composition of matter patent, a method-of-medical-treatment patent, several patents on specific formulations. . .when a generic company comes after you, these things are used as firewalls and are fought out as separate issues. And these would all have different R&D cost structures, although I've no idea how you'd figure some of them out, and thus presumably different patent lifetimes.
No, I think that passage of such a law would turn out to be the Cost Accountant and Patent Lawyer Full Employment Act of 200x. As I said, I can see the theoretical appeal. But in practice, I think that this attempt to reward the costs of innovation would only create more expensive makework, none of which would have anything to do with research itself. We've got too steady a supply of such as it is.
+ TrackBacks (0) | Category: Drug Prices | Patents and IP
January 28, 2004
The January 22 issue of Nature has a fine essay by Freeman Dyson (a hero of mine, I should add) about a fateful meeting he had with Enrico Fermi back in 1953. This was back when Dyson was a professor at Cornell, studying both the weak and strong nuclear forces.
There was a fine theoretical framework for the weak force (quantum electrodynamics, and a fine one it remains to this day.) But the strong force was giving people fits. Fermi was leading a team that did the first accurate measurements of the scattering of mesons by protons, the best data available on what the strong force was like. And after showing that QED did an excellent job on the weak force, Dyson had put his theoretical group to work in this trickier area.
After what he describes as "heroic efforts" (recall that these were the days long before any meaningful computing capacity), Dyson's team had a set of graphs of what the meson-proton interactions should look like, and they weren't too far off of Fermi's experimental data. So Dyson excitedly set up a meeting with Fermi in Chicago, showed him the graphs, and I'll let him take the story from there:
. . .he delivered his verdict in a quiet, even voice: "There are two ways of doing calculations in theoretical physics", he said. "One way, and this is the way I prefer, is to have a clear physical picture of the process that you are calculating. The other way is to have a precise and self-consistent mathematical formalism. You have neither."
. . ."To reach your calculated results, you had to introduce arbitrary cut-off procedures that are not based either on solid physics or on solid mathematics." In desperation, I asked Fermi whether he was not impressed by the agreement between our calculated numbers and his measured numbers. "How many arbitrary parameters did you use for your calculations?" I thought for a moment about our cut-off procedures and said "Four." He said "I remember my friend Johnny von Neumann used to say, with four parameters I can fit an elephant, and with five I can make him wiggle his trunk." With that, the conversation was over.
Dyson points out that, in hindsight, Fermi was absolutely correct. The theory they were trying to use could not possibly have done the job, not least because no one had a good idea of what protons were like (Gell-Mann hadn't come up with the concept of quarks.) Fermi, of course, was dead before quarks had ever been postulated, but he could tell that the existing framework was inadequate. And he saved Dyson years of what would have almost certainly been wasted time.
This is a perfect example of one of Weinberg's "Golden Lessons" that I spoke about on January 20th (below.) If you're working on a problem that no one (yet) has the power to solve, you can spend all your creativity in vain. Think of what would have happened if, say, Isaac Newton had stumbled across radioactivity. What could he have made of it? What are the odds that he would have been even close to correct? (And keep in mind, I'm saying these things about one of the greatest natural talents that science has ever known - Newton was downright terrifying.) A mark of a really great scientist, which Fermi certainly was, is to have a better eye for what problems are both significant and soluble. That's a small territory to work in sometimes.
In drug research, we work against a backdrop of doubts like this. Extraordinary new things are learned about living systems every year, and every time I find myself pitying all the people who were working on the same problem years ago. They may have suspected, but couldn't have know, what was really happening. Years from now, other scientists will pity us in turn. All the more reason to celebrate, when we actually get something to work!
+ TrackBacks (0) | Category: Who Discovers and Why
I wanted to mention another thing about Sanofi-Synthelabo's bid for Aventis. I hope that the French government keeps in mind, as they promote this deal, that one of the main consequences of such mergers is loss of jobs. Guys, that's the point. Associate directors of regional marketing, VPs for regulatory affairs - these people have to be thrown over the side, or the numbers just don't make sense.
In the nastier mergers, that crowd on the aft deck gets even thicker, and features R&D staff from the newly redundant therapeutic areas. Ideally, you'd want to hang on to those people (after all, this big new company is going to be doing more research, right?), but sometimes they get tossed. And that's not even taking into account many of the good ones who leave during the chaos for better (and, one hopes, more stable) new jobs - for now, I'll concentrate on involuntary departures.
Here's the question: where is this new French company going to cut jobs? Surely not in France! The unions there are famously fierce, for one thing, and it's hard to see how Chirac's government could be pushing so hard for something that will lead to thousands of itsconstituents being fired. France's unemployment is high enough already, thanks. So where?
Germany? That won't be popular. Die Arbeitslosigskeit problem is already the biggest issue in German politics. There are a lot of ex-Hoechst people in Aventis. . .but the Germans have already caught on:
The prime minister of the state of Hesse, Roland Koch, called on the German government to use its influence with French officials to block the unsolicited, $57 billion bid for Aventis by Sanofi-Synthebo, which he said could deprive Germany of jobs and access to research.
"The federal government has to bring its influence to bear against it," Koch said in a statement. "It's not just a question of exchanging shares, it's also about chances for jobs and research in the future." Germany's economics and labor minister, Wolfgang Clement, said Berlin would defend the company's jobs here.
So where? We're rapidly narrowing things down, unfortunately. As one of my correspondents put it, "I would be pretty nervous if I were in Bridgewater now." Bingo! That's the former Hoechst site in New Jersey, Aventis's biggest R&D in the US. It's not like they don't have a history of vacating buildings - try visiting what used to be the research sites of Marion Merrell Dow or Rhone-Poulenc Rohrer (that was a nice one, in its day.) I have to believe that this is where a lot of slack will be made up.
That'll be quite a feat. As was made clear in that Boehringer speech I quoted from last week, just about every other European pharma company is looking to expand operations in the US. And the new Sanofi-Synthelabo-Aventis might be forced to do the opposite, just to keep down the uproar at home. EU industrial policy at its best.
+ TrackBacks (0) | Category:
January 26, 2004
So Sanofi-Synthelabo is making a bid for Aventis. Shows you what I know about this business - if I'd guessed that this was in the works, which I didn't, I'd have bet that it would have been the other way around. Aventis would seem to be the company more in need of help. But as the New York Times article (by Andrew Ross Sorkin and Gardiner Harris) put it:
"Publicly, Sanofi-Sythelabo seems the picture of health. . .But almost every merger in the drug industry's long history has been found, in retrospect, to have been done because of undisclosed troubles that often left managers looking for outside help."
They have a point there. In this case, it's Plavix, for stroke, which is where Sanofi is coining most of their money these days. There's a patent challenge from a much smaller company, Apotex, and this sudden move has people wondering if Sanofi thinks that they might well lose the case. But what do they want with Aventis? That's the former Hoechst-Roussel-Marion Merrell Dow-Rohrer and a basket of others: a batch of mediocre companies who combined to make one really big mediocre company, as far as I can see.
Ah, but Aventis is headquartered in France. And this deal would create a mighty drug company, third largest in the world, and it would be French. This is just the sort of thing that the French government finds irresistable, and they've already waded into the fray by urging Aventis to just lie back and enjoy it while Sanofi makes a hostile offer for them. Try to imagine, say, the Bush adminstration's Commerce department doing that. . .
So is this going to work? That's actually a two-parter. First off, is the merger going to take place? I think that Chirac's government will force the issue, one way or another. There are lots of leverage points, since Aventis has several large stakeholders around the French industrial scene. Surely various people owe one another certain favors - after all, they all went to school together, n'est-ce pas?
The only thing that could derail things is that this is, in the end, an unsolicited offer. Why Sanofi-Synthelabo and Aventis didn't settle this in the traditional European fashion is a mystery. Other unsolicited offers could start coming in, assuming that someone has a burning desire to own Aventis (or a desire to see Sanofi not own them.) Novartis? Roche (but Novartis has a big stake in them, which might slow things down.) AstraZeneca? Bayer? I don't know if those folks have the money, frankly, even if they have the desire. Pfizer? It's been months since they bought someone; they're probably getting nervous twitches by now.
And the next question is, would the merger work? Probably no worse than others, but likely no better. The merged drug-company landscape is not a very inspiring one. A lot of people have been thinking that these deals disrupt more things than they resolve. It's possible that Sanofi is a bit behind the curve here, and even if they win, they could live to regret it. To quote the Times article again:
"Hostile takeovers are almost unheard of in the pharmaceutical business, because much of the value of drug companies is tied up in things that are both unpredictable and opaque to outside observers: the undisclosed results of clinical trials for new drugs, patents whose scientific and legal underpinnings are often closely guarded secrets and scientists who can leave. Buying a drug company without being able to peek inside its labs and legal files is extremely risky."
Well said! I couldn't have put it better even from inside the belly of the whale, here. Now, try an interesting thought experiment: read that paragraph again, and every time it says "drug company", insert the word "stock" after it. Hrrm. It makes just as much sense that way, folks. If that doesn't furrow your brow a bit, you must not own many drug stocks.
Update: Steven Jens points out that Sanofi's offer didn't exactly set Aventis stock on fire. . .
+ TrackBacks (0) | Category: Business and Markets
January 25, 2004
My experimental results came in late Friday afternoon, and. . .well, rarely have I seen less encouraging data. It wasn't enjoyable. I was there as the numbers for each part of the experiment came through, and I could tell early on that I was in trouble.
So, here's the rundown: the repeat of my previous (putatively successful) experiment failed. Then the attempts to increase the effect failed at each point - if anything, things went down instead of up. Then the attempt to reverse the effect (in four different ways) failed, no doubt because by this time there really wasn't anything to reverse. The best-looking run of the whole afternoon was from this group - but it was, perversely enough, the one that should have been the most shut down. Ungood.
What now? The only way this experiment can be any kind of good news is if some systematic error disabled the whole thing. I would love to find out that one of the components was taken from the wrong vial, or was left on top of a hot plate or something. But that's highly unlikely. But what about that previous experiment, the one that led to this death-or-glory attempt? I'm going to be going back over that one, giving it a fishy glance in light of what happened on Friday. At this point, the hypothesis that best fits the data is that the encouraging results are wrong.
I may still be able to do one more attempt. There are a couple of oddities about this latest data set that I don't understand, and the idea is worth one more shot under the cleanest conditions I can think of. But that, for now, will have to be it. I'm going to have to go back to the drawing board and think about what's going wrong, see if there's some different way to realize what I still think of as a beautiful idea. There may yet be. Plenty of beautiful ideas don't work, though: beauty is necessary, but not sufficient.
Science is fun, it really is. And it's certainly damned useful. But it isn't easy.
+ TrackBacks (0) | Category: Birth of an Idea
January 22, 2004
I spoke a couple of weeks ago about my latest series of experiments at work, and I've had several inquiries about how things are going. Well, the whole shebang has been in the freezer, actually. The instrument that we need to analyze things (and the person who runs the instrument!) have both been occupied with an unexpectedly lengthy problem in one of our drug discovery programs. My high-risk side project takes a back seat to that, understandably.
But the thaw is coming. I was told yesterday that my samples have moved back to the top of the list, so it's possible that I'll start to get results tomorrow (and if not then, on Monday.) So, here we go. I think that these runs are going to be definitive, one way or another. There's always the possibility of a "maybe" answer when you do an experiment, but the key to successful design is setting things up so you don't get many of those. I think these results won't have too much ambiguity in them - they shouldn't. If they do come back fuzzy, it means that my mental picture of what's going on is faulty, even if my broad idea turns out to be right.
My whole idea is on the chopping block, and the knife is poised to fall. I'm going to be able to look back on this evening, this whole period, as either the last time when I still had doubts about this discovery - or as the last time when I still had hopes that it was real. Oh, science is fun - it really is. And it's certainly damned useful. But it isn't easy.
+ TrackBacks (0) | Category: Birth of an Idea
January 21, 2004
Genetic Engineering News reprints parts of a speech given by Rolf Krebs, chairman of the German drug firm Boehringer Ingleheim, at a recent conference in Hannover. Dr. Krebs was speaking on the differences in pharma research between Europe and America, and he didn't leave much bottled up:
"The framework conditions for the pharmaceutical industry in Europe and the U.S. could not be more different. . .Germany serves as a good example of these changes. Last year, the "pharmacy of the world," as Germany used to be known, reported its first-ever negative export-import ratio. In the meantime, the government has realized that of the seven pharmaceutical firms previously doing R&D work in Germany, only four remain. . .it is alarming that only ten of the world's 185 industry research centers are now located in Germany. This is symptomatic of the situation throughout Europe, where 20 research centers have been closed down over the past six years."
All true. It's been clear for some time now that pharma and biotech companies are not expanding in Europe - the American ones aren't going there, and the European ones are coming here. After going over the well-known problems the industry has been having with fewer good drug targets and more complicated drug development, Krebs says, correctly that this situation raises the risks for all existing project. The higher financial stakes make late-stage project decisions very important, and those decisions are based (naturally) on expected economic return. And where do you go for that return?
"It is becoming increasingly evident that the success of the major pharmaceutical companies depends on the extent of their U.S. presence. . .There are many reasons why the U.S. market has acquired greater importance compared with the European market. Although growth has slowed more slightly over the past two years, the U.S. market is still expanding much more dynamically than the European market."
And what brings on such differences? The answer "money" is always an appropriate first guess in these situations, but there are several factors at work:
"In contrast to the high degree of freedom that exists in the U.S., all the individual European markets are subject to state regimentation. This applies both to pricing, as well as to the recognition and reimbursement of pharmaceutical products by the statutory health insurance funds. This lack of regulatory intervention in the U.S. has had a number of positive effects for the pharmaceutical industry. Prices are based on therapeutic quality; products are subject to a cost-benefit analysis all along the line (including patients.)"
This is a bit idealized, but compared to the European situation, he has a point. Krebs goes on to contrast the fundamentally different attitudes that have led to this situation:
"In comparison with Americans, we Europeans are less innovative and, above all, less inclined to take risks. We take comfort in the argument that it takes longer to create the conditions for new technologies due to our democratic processes. But democracy as such is not the root cause of these regrettable delays, as evidenced in the U.S. and the U.K., both of which are democratic states. The unwillingness to innovate and take risks results directly in an overabundance of rules and regulations. We want to lay down the final result without gaining the necessary experience first."
As he goes on to say, it's harder to get venture capital in Europe, and if you do manage to get a new idea off the ground, you run into a mass of inconsistent regulation. Hack your way through that, and your reward is compulsory government pricing. "The message is clear," he concludes, "it is an advantage not to invest in innovation."
Having observed both the American and European pharma industries at close range, I think there's a lot of truth in what Krebs says. It's not that the European companies haven't done good work, just that they could have done still better in a different environment. Being intelligent and capable, they've come to understand their problem. And finding themselves with nowhere to go in their home countries, the French, German, and Swiss firms make their cutbacks at home - but not here. Or if they're in better shape they break ground on their new research centers - not in Basel, but in Cambridge. How much of a company's R&D has to take place outside Europe before they're not a European company any more?
+ TrackBacks (0) | Category: Business and Markets | Who Discovers and Why
January 20, 2004
Nobel laureate Steven Weinberg had a piece back in the Nov. 27 Nature (p. 389) offering advice to people just starting their scientific careers. It's useful stuff, and the lessons aren't just for beginners, either.
His first of "Four Golden Lessons" is No one knows everything, and you don't have to. (This came from his early paralysis at not knowing the whole field of physics.) That one gets more true every year, as the pile of scientific knowledge increases. I'm a reasonably good organic chemist, but there are big swaths of the literature that I'm not familiar with. Unusual steroid reactions? They're legion, but I don't know 'em. Sesquiterpenoid biosynthesis? (Half my readership just said "Gesundheit!") They're wild-looking compounds (PDF), but I've never needed to know much about how they're made. Mechanistic organometallic complex chemistry? Go ask Greg Hlatky (and while you're there, check out his piece on the compounds these things can make by reacting with stopcock grease).
No, I don't know these things too well, and (so far) I haven't had to. If I need to, I'll learn them. That's the best way to deal with the size of a scientific field, I think: get the fundamentals down, and that will give you the tools to learn what you need to know. Then let your own curiosity and your circumstances take you from there.
Weinberg's second lesson is aim for rough water. Try to work in a field where things are contentious and unsettled - "go for the messes." There's still room for creativity there, as opposed to the more worked-out fields. Of course, that presupposes that the reader is interested in doing creative work, but advice like this is wasted on anyone who isn't. Not that there aren't plenty of such people around - any research department is full of them. They can make contributions, as long as both they and their supervisors know the score. Trouble ensues when they don't.
His third lesson is forgive yourself for wasting time. He classifies that as "probably the hardest to take" of his lessons. This is a consequence of the go-for-the-messes advice, and will be most applicable to those that have followed it. What he means is that it can be very hard to know if you're working on something that's even solvable, or if you're working on the right problem at all.
That certainly applies to my area of research, where there are long stretches where it seems like nothing's happening, and perhaps never will. Even when things are moving along, you're never sure if that light off in the distance is reflecting off a pile of gold, or is a bare bulb put there to scare the rats off a garbage pit. So, we're searching for an agonist of receptor XYZ - who knows if such a thing exists? Or if we're going to stumble across it? Or if it'll do what we think it will, assuming we know how to test it in the correct way and assuming that we can understand the results? Working like this, there's going to be a lot of wasted effort, and you're just going to have to come to terms with it.
Weinberg's final lesson is learn something about the history of science. The least important reason to do that, he says, is that it might help out your research. To use his example, without knowing the historical record, you might come to believe that Thomas Kuhn or Karl Popper really understood how science works. But the larger reason is that an appreciation for history puts your work in perspective. Weinberg believes, as I do, that science is one of the highest activities of civilization, and that we should be proud of our parts in it. A real discovery can live longer, and with greater impact, than almost any other human work.
+ TrackBacks (0) | Category: Who Discovers and Why
January 19, 2004
Signaling between cells is weirder than we used to think it was. There's a hardy perennial, all right - that sentence could have been written whenever you like for the past fifty years or so. But the surprises keep on coming. Some of the most intense communication needs are between neurons, as you'd expect, and it looks like nature has taken advantage of all kinds of things to achieve greater bandwidth.
Everyone now learns about nitric oxide and its effects when they study physiology. The thought of a toxic gas as a neurotransmitter was a tough one to deal with at first, but the evidence was overwhelming. Then in the 1990s, two more oddities were proposed in the same category, and they're even more poisonous: hydrogen sulfide and carbon monoxide. Comments of the "You have to be kidding" sort greeted the initial work in this field, but the nitric oxide work had opened the door. It now appears that these two are, indeed, important signaling molecules in the brain.
Hydogen sulfide has a number of physiological effects, other than poisoning you (or, in lower concentrations, making you choke on its delightful aroma.) It seems to act on smooth muscle along with nitric oxide (now, there's a combination I would go out of my way to avoid breathing), and it also seems to have a role in laying down long-term memory. Carbon monoxide also seems to have a number of different functions - it's vasoactive, like its gaseous partners, but also seems to be involved in the immune response and in cellular protection and repair.
Taking up in the same way as the nitric oxide research, which has stimulated a huge amount of drug discovery work over the years, people are now trying similar tricks with these new gases. Look for more and more work on these in the drug industry as their mechanisms get fleshed out.
What's next? Well, it's not impossible that some other small-molecule gases have their own pathways, too. These things have properties that aren't shared by any other molecules, and perhaps they're being put to use. Ammonia, sulfur dioxide, and nitrous oxide have all been proposed as candidates. Thinking along those lines, I have to wonder about the small alkyl derivatives like methylamine and dimethyl sufide, too. Why not? But if someone gets around to claiming chlorine or the other halogens, I'm going to start to wonder. And if there turns out to be a physiological role for the noble gases, I'd start to suspect that Einstein was wrong: maybe God's approach to scientific laws is malicious after all. It would explain a lot, now that I think about it. . .
+ TrackBacks (0) | Category: Biological News
January 18, 2004
Remember the genomics gold rush? Back about five or six years ago? Sure you do! People were lining up to throw money at companies that could deliver human gene sequences, as part of the never-ending search for new drug targets. (OK, it's not quite never-ending, but for the time horizon we have in the industry, we'll stick with that adjective.) Well, a good part of the reasoning behind all those sequencing deals may have just taken another hit.
Even while the genomic craze was at its peak, there were doubters. Gene sequences should, in principle, read out directly into protein sequences. But there are already some complications, since DNA and RNA sequences can, at various points, be spliced and recombined. We think we know the signs of that happening, but it's still another thing to worry about.
But even if there's no funny business, it's not easy getting useful information from a raw sequence. We still can't predict protein structures de novo, not to the degree that medicinal chemistry needs. You can learn a few things - homology statistics can place your unknown protein into a known family, or (failing that) at least tell you stuff like whether it's likely to be membrane-bound. But knowing that something's, say, a G-protein-coupled receptor sure isn't enough to tell you what it does in vivo, or if it's a valid drug target (and if so, for what disease.)
And there's always been a lot more to the cohort of proteins than just the corresponding genomic sequences. That's where the doubting voices got louder. Proteins get modified in all kinds of ways. They get phosphorylated and glycosylated around their outsides, for example, which can profoundly change their function. And they can get sliced up into smaller proteins, too. Happens all the time - plenty of bioactive proteins are produced from a larger percursor, carved off as needed like sandwich meat at the deli counter. (The enzymes that do the carving can be very good drug targets indeed.)
Enter the latest craziness, from J. C. Yang's lab at the National Cancer Institute. There's an exhilarating (or alarming, depending on your point of view) paper in the ">latest issue of Nature (427, 252), whose authors have seen something that no one had ever seen in higher organisms. They've shown that not only can proteins be chopped up in the cell, but that the various fragments can be spliced back together in new combinations. In their case, they showed that cells could produce a nine-amino-acid peptide from a 49-amino-acid precursor. The middle 40 got snipped out, and the two ends were spliced together to make the nine-mer. You're never going to be able to read off the sequence for that one, now, are you?
This sort of thing goes on all the time in single-celled creatures, and is known all the way up to, oh, bean plants. But it had sure never been seen in mammalian cells. How does this process happen, and how important is it? Who knows! It might turn out to be a rare curiosity, or it might turn out to be something really important that we've completely missed seeing all these years. To be sure, no one's reporting coming across a lot of important proteins whose sequences couldn't be matched in the genome somewhere. But there are an awful lot of proteins whose sequence we don't know, so the upper and lower bounds of this new phenomenon are fuzzy.
This paper, you can bet, has already set off a flurry of research. Perhaps there are some unexplained proteomic problems out there which this will turn out to answer. And here's a prediction: it wouldn't surprise me if protein splicing had already been seen by someone else, who looked at the data, thought about it, and said "Naaaah. That can't happen. Must have messed something up somewhere. . ." If this turns out to be physiologically important, it's Nobel material for sure. Listen closely, and you may hear the sound of someone kicking themselves.
+ TrackBacks (0) | Category: Biological News
January 15, 2004
Here's something new to finish off the week. As even casual observers know, patents are extremely important in the drug industry. We spend whacking amounts of time writing them, reading them, and worrying about the ones that other companies have filed. And it should also be clear, to the same laid-back observers, that a good number of patents are. . .well. . .not what they could be. Many are a bit. . .overstated, or perhaps a bit under-enabled. Estimates vary. Some people think it's 90% of all patents. Others more conservatively estimate that only half of them are nonsense.
It's taken a team effort for things to come to this. Patent applicants have to write crummy applications; the patent office then does its part by granting way too many of them. And that leads up to this latest bolt of inspiration, whose examination by the PTO I would pay to witness. Prepare yourselves for US application 200400055535: "Process of Reincarnation."
Uh-huh. Allow me to quote from the Summary of the Invention. Heck, let's do the whole thing; it won't take long:
" This invention resulted from my combining Einstein's Theory of Relativity and Newton's Second Law of Physics." (sic)
" Reincarnation is defined in Webster's Third New Inernational (sic) Dictionary as "rebith" (sic). Thus my invention is a process of rebirth or in other words immortality."
There we go. Actually, I could have saved myself some time by just putting (sic) after the whole damn thing. You can guess what the single claim consists of. It's the shortest patent application I've ever seen, which is only one of its many distinctive qualities. But - and I'd hate to break this to the applicant in person - I don't think it has a chance.
You see, the, er, inventor didn't cite all the Buddhist and Hindu prior art. It's the sure sign of an intellectual-property amateur. If he'd just worked in a couple of references to the Tibetan Book of the Dead, say, I think he could have been out the door in a matter of weeks with his issued patent. But as it is, it'll probably take months for approval. Details, details.
+ TrackBacks (0) | Category: Patents and IP
January 14, 2004
I've been meaning to get around to the subject of Abbott's HIV protease inhibitor, Norvir (ritonavir.) Actually, the subject I've really need to get around to is its price, which in December went up by about a factor of four. That's a pretty steep move for something that's been on the market for seven years, isn't it?
Norvir is a protease inhibitor, one of many these days. It was a front-line therapy when it came out, but now it's settled into an unusual role as a supporting player. When added into other "cocktail" treatment regimens for HIV, Norvir seems to make everything more potent. There's no voodoo involved here, actually. Norvir turns out to be a good inhibitor of one of the liver enzymes (CYP3A) that is responsible for metabolizing many other medications.
That effect is enough to stop development of some drugs, to tell you the truth. No one is going to take something for a lesser illness that's a CYP3A inhibitor, because it'll make the blood levels of so many other drugs go haywire. But for HIV, especially when Norvir first came out, all sorts of side effects were tolerated. And over the years, the liver enzyme inhibition has turned from a bug into a feature. (It's still something to look out for, though, if a patient taking Norvir also takes things like sidenafil (Viagra) or a statin - see this PDF for details.)
But there's a side effect of the side effect. Norvir's taken at lower doses than other anti-HIV drugs, because it's only partially being given for its intrinsic protease inhibition. Back in 1996, as a stand-alone therapy, it was given at 1200 mg/day, which cost about $20. Today, as an adjunct to other therapies, it's given at about 100 mg/day, with a corresponding decrease in Abbott's financial expectations.
Thus the price hike, or so one has to figure. Abbott has said that the new price "better reflects the current market value of Norvir," and they point out that newer therapies run in the $20/day range, while Norvir, even at the new price, will be $8.57/day, at a 100 mg dose. Of course, that would make the original 1200 mg/day dose pricey indeed, but you don't see Abbott mentioning that. (Here's their letter to physicians (PDF) on the issue.)
Companies can charge what they want to for their products, or at least they should be able to. So Abbott is completely within its rights to raise their price to try to recoup some of what they originally thought Norvir might bring in, although you wonder if the bad press doesn't cancel that out a bit. But there's another explanation that looks even less appealing: Abbott has a more recent combination HIV therapy (Kaletra) that has Norvir as one of its components.
The new Norvir price now makes other combination therapies more expensive than Abbott's own Kaletra. So in their way, they may well be trying to compete on price here, but by raising the price of their competition rather than lowering their own. That's a rough way to do business, but given the state the drug industry's in, it wouldn't surprise me if that's exactly the plan. Profits are where you find them. I'm sure that Abbott's competitors have already made up their minds about what's going on.
So, is it right to do this sort of thing? This is where arguments about drug prices diverge from arguments about, say, the price of pearls, chocolate chip cookies, or ski equipment. There's an unavoidable moral aspect that comes into the discussion when you're talking about patients with a deadly disease. It wouldn't be nearly as big an issue if Pfizer, say, racked up the price of something that hypothetically made Viagra work better. But HIV's different.
On the purely pragmatic plane, I realize that I just said up there that I think that companies should be able to charge what they want to charge. I'll stick to that, but that doesn't mean that it's always a good idea for them to go ahead and do it. Abbott needs to realize that its actions look - no matter how strenuously they deny it - like an attempt to body-check its competition by making everything more expensive. And they need to realize that we pharmaceutical companies might not have the pricing power that we once did. If we keep doing this kind of thing, we're going to end up with no pricing power at all, and people will be clapping and cheering while we all tip over into the trash. Think about it, guys.
+ TrackBacks (0) | Category: Drug Prices
January 13, 2004
I wanted to take a moment to mention some interesting posts around Blogdom that readers may not have seen. In a response to the news on secretin for autism (see my post below), Dwight Meredith writes on what it was like at its peak of interest:
Human secretin, swine secretin, herbal secretin (which as far as I can tell is an oxymoron) and synthetic secretin were all hawked relentlessly to the parents of autistic children. The price of secretin skyrocketed. People were paying $2,000 for an amount of secretin that before the buzz had cost about $30. It is not an exaggeration to say that parents were mortgaging their homes to purchase secretin for their kids. We now know that a sugar pill would have been equally effective.
Please note that all of that buzz was generated by the fact that a few autistic children had improved after being given secretin for digestive problems. The autism community could not wait for double blind and placebo tested trials. We wanted our miracle and we wanted it now.
This is a man who writes from personal experience, I should note. And I can understand the desperation (well, as much as anyone in my position can - I have two small children, neither of whom have - thus far - shown any neurological abnormalities.) What I have trouble imagining, though, is what goes through the mind of someone who peddles "herbal secretin" to parents who are begging for something to help their autistic child.
Herbal secretin? They didn't even bother making it sound like anything but a heartless scam. Figured the customer base would be too desperate to care, I suppose. I'm ashamed to be in the same phylum with creatures who would do something like this.
There's a larger point about the wait for double-blinded trials, too, of course, which I should save for a longer post. The short form is that I can see the point that some people make, that it would be better to require safety (Phase I) trials, then stand back and let efficacy be sorted out in the marketplace. (SMU's Steve Postrel and I had a long e-mail exchange on that subject a year or so ago.) But then I hear about this sort of thing, and start to think that this is one of those sensible ideas that would only work on some other species than humans.
The other post I wanted to mention is over at Colby Cosh's site. Talking about medical progress, he hits on the idea of looking at the causes of death in the records of ballplayers from the old days, who were in their physical prime. It's an alarming list, and most of the things on it are, fortunately, in the process of disappearing from the world. And good riddance. As Cosh says: "I don't know how anybody kept from just going insane before antibiotics existed, with death lurking around every corner."
One final note - I've forgotten to mention that Charles Murtaugh is back blogging again. There's lots of good new stuff; just start at the top and work your way down.
+ TrackBacks (0) | Category: Autism | Blog Housekeeping | Clinical Trials | Drug Industry History
January 12, 2004
Time for an update on my research, where I'm still working on the odd idea that I've been speaking about. In my last installment, I had what seemed to be good results from an experiment, and I was getting ready to set up some more control runs to see if things would behave as they should.
Well, those reactions are going right now, thanks to the efforts of a colleague in Biology. These experiments will be running all night (Monday EST) and finish off around lunchtime tomorrow. Then I'll take the solutions down to another colleague in our analytical department, and in the next few days, she'll tell me what's happened. The wait will not be an easy one. I can tell that already.
That's because this batch of experiments is actually a pretty strenuous interrogation. I've tried to set it up so that good results can come out of it only if there's something real going on. I think there is, of course, but it's impossible to say for sure. My results from the first experiments could be characterized as "consistent with my hypothesis," but that's all. Mind you, that's a lot better than the alternative, hoo boy, but there could be other (less interesting!) things they're consistent with.
But the reactions taking place tonight should sort things out, but good. This run has four different parts to it: There's a repeat of the most promising conditions from the first experiment, just to ask the most basic question (reproducibility.) A distressing number of interesting experimental results never poke their heads up again, so that's one hurdle. In the second part, there's a set of conditions that should cause a larger effect than I saw the first time. This attempt is being racheted up in two seperate steps. If it goes up nicely each time, I'll be very happy. If the results come back one-up, one-down, I'll be staring out the window a lot, trying to figure that out. And if they show no effect, well. . .
The third part reverses field: it's an attempt to completely abolish the effect, by a mechanism that should be quite specific to my hypothesis. This one's in two steps, too, in another attempt to see a dose-response relationship. Having this one come through, which would revert the system to the same results coming from the corresponding blank experiment, would be strong evidence that I'm on the right track. The reverse holds true, too, unfortunately - if there's no effect here, my hypothesis has taken a torpedo right in its engine room. (That blank experiment is running tonight, too; it's an important control for all these tests.)
And the last ring of this circus is another attempt to make my desired effect disappear. I've changed a chemical structure in a way that should make very little difference to anything, except in the case of my hoped-for mechanism. It should shut that down pretty cleanly. It'll be hard to hold on to my current idea if this doesn't work as planned, either. I'll have to fall back on experimental error, which is not the first explanation you want to reach for, or some other variable that has completely escaped my notice. Neither of those is a good bet at this point; it'd be a lot simpler to assume that nothing interesting is happening at all.
For readers outside the research arena, those try-to-kill-it experiments are a powerful and commonly used technique. It's hard to run them sometimes, because it's hard to escape the mental picture of your new phenomenon, just arrived into the world of physical experience, being scared back into its hole by the sudden advent of search lights and sirens. But, you know, there are a lot of things to work on in this world. And if you don't figure out what's real and what isn't, you can spend most of your scientific career doing the equivalent of digging holes and filling them back in. It's hard on a hypothesis, being put to the test like this, and I'm here to tell you that it's not all that easy on the person behind the idea, either. If something's real, though, it'll show itself - it'll have to show itself - no matter what nasty questions you ask. Better to ask them up front.
+ TrackBacks (0) | Category: Birth of an Idea
January 11, 2004
In my industry, you hear a lot of talk about drug targets and their relative chances of success. Targets fall into several broad classes, and when you take a close look, there are clearly some that are easier to hit than others. The G-protein coupled receptors (GPCRs) are one of those (antihistamines and beta-blockers are classic examples), and various hydrolytic enzymes are another (ACE inhibitors, HIV protease inhibitors, PDE inhibitors like Viagra, etc.)
But there are some other categories that are severely under-represented. "Interaction" targets is what I'd call a broad group of these. The ligands for the easier enzymes and GPCRs fit into defined binding pockets, which have evolved for small molecules, It's the old lock-and-key picture. But trying to affect the binding between two proteins, or of a protein with a stretch of DNA/RNA - now, that's something else again. There's no single binding pocket there, at least not on the scale of a drug-sized molecule. Instead of fitting different-shaped keys into existing locks, we're faced with trying to wedge something in between a door and its frame.
It's hard to get in there, and our molecules are often too small to have much effect. But the number of drug targets in this class is huge; we're going to have to come to terms with them eventually. . But for now, one of the best ways is to carefully study the various high-value targets and see if there are some that look more likely to work, given what we already know how to do. That's what a group at Roche has been up to recently, and they've reported their success in an online preprint in Science.
They're after a protein called MDM2, which acts as a brake on the activity of a more famous protein from the p53 tumor-suppression gene. In many cancers, it would be good to block this interaction and get the p53 system as back to being revved-up as possible. (Of course, in many other cancers, this gene has already been taken out of action by one mutation or another, which is probably a key step in their formation. Those won't be candidates for MDM2 blocking therapies.)
In 1996, a group at Sloan-Kettering published an X-ray crystal structure of the two proteins, which showed that there was a fairly clear pocket that seemed responsible for a lot of the binding. It looked like a possible candidate for a small molecule, but this is the first report of real success in targeting it (although others are hard at work.) The Roche group found some polyaryl imidazoline structures through high-throughput screening that seem to do the job. One of them is even orally active in a rodent tumor model, which is quite an accomplishment. And as proof of the mechanism, the compounds are inactive against those cancer cell lines that have already lost their p53 gene.
This is good news, since we can always use another route to cancer therapy. But I'm not sure how broadly applicable this is going to be. I'm sure that there will be talk of new interest in protein-protein drug targets, but this one is (unfortunately) an anomaly. That type of small, reasonably well-defined pocket that plays a role here doesn't show up that often, and it's not like people haven't been looking. News that these things can succeed will stimulate more work in the area, true. But that's where a lot of the effort was going already, because other protein-protein targets have seemed destined to fail.
My mental picture of those targets is of two oil tankers slowly coming together, brought closer as dozens of small grappling hooks whiz out and clang onto different parts of their decks. With a small molecule, we're trying to interfere with that by sticking a fishing boat in between them. Not easy, but we're going to have to figure it out eventually. Protein-protein interactions are a hot topic these days (go off and Google "proteomics", but stand well clear while you do it!) so we're bound to learn a lot more in the next few years.
For now, congratulations to Roche as they move forward toward the clinic. They'll be the first to find out what blocking MDM2 binding is going to do to animals - how well it'll treat those with cancer, and what side effects it might have on those without. I hope there's daylight in between those two groups!
+ TrackBacks (0) | Category: Biological News | Cancer
January 8, 2004
A fellow researcher, working over at The Competition, sent along a couple of good questions. He's a biologist, and was reading the posts here earlier this week about compound repositories. He writes:
"I would like to pose to you a question I have tried for years to get chemists to address: why has no one yet come up with a better "universal" solvent than DMSO for dissolving compounds for routine screening in assays? As you point out, DMSO has numerous liabilities for this purpose (perhaps more on the biological side), and yet we all continue to use it routinely, because there doesn't seem to be a better alternative. In the various screening labs in which I have been involved for the past dozen or so years, we have tested a number of possible alternative solvents on an ad hoc basis, none of which seemed any better. Surely modern organic chemistry can do better?"
All I can offer my colleague is this: modern organic chemistry may not be quite as powerful as you've been led to believe. Perhaps the problem is that you've been listening to too many of us modern organic chemists. We do tend to go on.
He's right that DMSO has its down side, and there are some that I didn't even mention. For one, DMSO and air make for a decent oxidizing system, enough to cause trouble in electron-rich molecules. Things will start to change color on you in a DMSO solution that's been left open. And it does have its biological problems. Too much DMSO in an assay system will cause the proteins involved to change their conformations, probably inactivating them. At the very least, your data start to go haywire.
Is there anything we mighty chemists can do about this? Well. . .actually. . .no, probably not a whole lot. The problem is, anything that has "universal solvent" properties is going to have "universal denaturant" properties when it comes to large biomolecules. Proteins, carbohydrates, and nucleic acids are made to hang around with water. Anything that isn't water is going to cause trouble, sooner or later.
Coming up with a solvent that acts just like water, but isn't, may well be impossible. Water's just too weird. Its boiling point and viscosity are way off the estimates you'd get from looking at related things like ammonia and hydrogen sulfide, due to its extraordinary hydrogen-bonding powers. And it's those bonds that do the trick with biomolecules. It's sui generis - there's no other molecule that small that can do hydrogen bonds that strongly, at those angles, in two directions at once.
DMSO gets by because it's also small (although not as small as water.) It's the smallest sulfoxide possible, so it has the most character. The key is that the sulfur and oxygen atoms in the sulfoxide bond have a lot of charge on them - the oxygen's nearly a minus charge; the sulfur's nearly a plus. That dipole lets it really line up with any polar groups a molecule might have, and its two methyl groups give it a chance to dissolve some hydrocarbons that water won't accept. (Larger sulfoxide analogs just add greasiness, and are less powerful solvents. That's the wrong direction, and you can't go any further the other way.)
And all the other attempts at DMSO substitutes tend to follow that same path, things that are polar because of their high dipole moments. Some of the also-rans are N-methylpyrrolidone (NMP), DMPU, and the toxic HMPA. None of them are as good as DMSO, and they all suffer from its disadvantages. There may well be some funky structures out there that haven't been given a fair hearing, but I sure can't think of many right now. I'm afraid that we're just going to have to live with DMSO, and respect water's magical powers for what they are.
+ TrackBacks (0) | Category: Drug Assays | General Scientific News
January 7, 2004
Regular readers will know that I often rant about pharmaceutical price controls. The issue of Canadian drug reimportation is what usually sets me off, but that's far from the only border where pharmaceuticals can be profitably arbitraged. Take the countries of the EU, for example. There's no central EU health plan (not yet, anyway - can you imagine what an empire-sized bureaucracy that would be?) Each country's national health plan negotiates its own price. And some of them are indeed lower than others (Spain, for example.)
So, what happens is just what you'd expect would happen. People try to buy in the cheap places and sell in the more expensive ones, which activity has probably been going on since the goods involved were spear points and mastadon tusks. The middlemen are happy; the drug companies aren't. The only way they've been able to do anything about it is to try to restrict the amount of wholesale stock that makes it to the lower-priced markets, which is the same thing that we now see happening here with the Canadian pharmacies.
Bayer got dragged into court a few years ago for doing just that, and they lost their case on antitrust grounds. Fines ensued, and Bayer appealed to a higher court, which ruled in their favor in 2000. But another appeal went through to the European Court of Justice, which has just ruled in Bayer's favor again, to the delight of the pharma industry. Bayer's comment was particularly pithy: they expressed relief that they "are under no obligation to supply the entire European market from the member state with the lowest state-regulated prices." Just so.
It's not not just the pharma industry that applauded this news - car prices are just as out of whack across Europe, for example, and the same sort of games are played. As you'd expect, I agree with the court's decision, too. I'd much rather live in a world without price controls at all, but if we're going to let governments restrict the price of goods, then I think we should give the producers the option to restrict their supply. Trying to have it otherwise is like legislating sunny weather and free ice cream, which would be vote-getters, too, come to think of it.
+ TrackBacks (0) | Category: Drug Prices
Secretin is a signaling peptide that stimulates the pancreas. It kicks in when food is consumed, causing the pancreas to secrete its mixture of digestive enzymes. Until 1996, not many people (outside of gastrointestinal specialists and researchers) had heard of it, or devoted much thought to it. But that year, an autistic boy from New Hampshire, Parker Beck, was given secretin to help diagnose some GI problems - and his autistic symptoms apparently improved.
Well, that touched off some interest, as you can imagine. There is no generally accepted therapy for autism; the demand for one is terribly acute. There was a complex patent scuffle involving the University of Maryland and the Beck family, and Mrs. Beck ended up listed (quite properly) as an inventor and (more unusually) the assignee of the rights to the resulting patent for secretin as a treatment for autism.. She licensed the patent to a small biotech, Repligen.
A secretin underground sprang up very quickly, as you'd expect. Parents were trying it all over the place, and anecdotal evidence was all over the place, too: it worked, it didn't work, it worked a bit, it started to work, it stopped working. Controlled clinical trials were the only thing that could sort out the confusion. (As a drug company scientist, you'd expect me to say that, but I say it because I'm a scientist, not because I work for a drug company.)
I should emphasize that secretin as an autism therapy is not at all a crazy idea. First off, just about anything that shows possible efficacy for a terrible untreated condition is worth a look. Second, many autistic patients show gastrointestinal abnormalities, for reasons that aren't clear at all. (Are they related to the cause of autism, or are they an effect of it? Or are they just a distant sideshow? No one knows.) And secretin is found inside the brain (like a dizzying array of other peptide hormones), so it could well be a player in brain development. There's constant crosstalk between the brain and the digestive system, since eating is a rather important physiological response. The systems involved are complex and only partially understood (just ask anyone who does research on obesity and eating disorders); all sorts of things are plausible.
But interesting hypotheses are fragile things. The early trials of secretin didn't look very promising, but those were small samples. Further trials didn't look very convincing, either. Repligen's multiple-dose Phase II trial didn't do much better. By this time, many practitioners were abandoning hope. There have been a number of possible therapies for untreatable conditions that have looked promising anecdotally, but didn't hold up to scrutiny, and secretin was looking like another addition to that sad list.
Repligen pressed ahead to Phase III trials, hoping that a larger number of patients would show enough of a statistical effect to gain FDA approval. The first of these results have come out, and that's what was making the headlines this week. This is the largest trial of secretin so far, and it appears, again, almost completely negative. There's a hint of efficacy among a subset of higher-functioning autistic patients, but any time you hear that kind of post facto reasoning you should stay in your seat. Statistical power starts evaporating when you so subset analysis - you actually need more convincing evidence when you slice and dice the data like that, not hints of something that might be happening. That sort of thing is almost always noise.
So, as best as can be determined, secretin is not a viable therapy for autism. There are still autistic patients who appear to have been helped by it, although by this time you have to wonder how many of these cases are just due to chance (and how many are due to understandable observer bias.) Autism is a very heterogeneous disease, though, which makes it still possible that in some cases secretin is helpful. But we don't know how to diagnose these - if anyone figures that out, a further clinical trial would certainly be warranted. For now, though, secretin does not seem to be of any general use.
But that probably won't be the end of the story. I expect that there will still be desperate parents who will try secretin. And there will probably be some who will believe that it's helped. And as long as there's a market, there will surely be people who will supply secretin, at a cost of many thousands of dollars a year, and whether it does any good or not isn't going to be a concern of theirs at all. A sad situation.
+ TrackBacks (0) | Category: