Corante

About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png 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: derekb.lowe@gmail.com Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Emolecules
ChemSpider
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
PubChem
Not Voodoo
DailyMed
Druglib
Clinicaltrials.gov

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
Kilomentor
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
ChemBark
Realizations in Biostatistics
Chemjobber
Pharmalot
ChemSpider Blog
Pharmagossip
Med-Chemist
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
SimBioSys
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Business|Bytes|Genes|Molecules
Eye on FDA
Chemical Forums
Depth-First
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa


Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
FuturePundit
Aetiology
Gene Expression (I)
Gene Expression (II)
Sciencebase
Pharyngula
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net


Medical Blogs
DB's Medical Rants
Science-Based Medicine
GruntDoc
Respectful Insolence
Diabetes Mine


Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem


Politics / Current Events
Virginia Postrel
Instapundit
Belmont Club
Mickey Kaus


Belles Lettres
Uncouth Reflections
Arts and Letters Daily

In the Pipeline

Monthly Archives

February 29, 2004

More From the Me-Too Front

Email This Entry

Posted by Derek

Some interesting mail has come in after last week's post on comparative clinical trials. Reader C.B. that I spoke about here some time ago, but should have raised again:

"It seems to me that something else is being left out: not all patients respond the same way to any particular drug. . . Suppose that drugs X and Y are equally efficacious when given to the appropriate patient, but the population more responsive to X is smaller than that benefiting from Y. A simple comparative trial would suggest that Y was more effective because it assumes a single type of patient. On the basis of the results, people who should get X would only be allowed Y. . ."


It's true, there are a number of cases like this, and this is one of the traditional arguments for multiple drugs in a given class. I've made it myself. Given the state of the art, it's nearly impossible to untangle these things. In almost all cases, we have no idea why some people respond better to a particular therapy; it's trial and error. Clinically, these things are bottomless pits, so I think that comparative trials are going to be most useful in areas where a large number of patients respond to both drugs under study.


But we're in the process of inventing ourselves out of this situation. That's why all that money is being poured into pharmacogenomics - and quite rightly, although the end result is that many drugs are going to have their potential market size whacked into a rather more compact shape. The great thing about pharmacogenomics is that we're finally going to know who should take our latest drug, and we'll be able to find them and sell it to them. The terrifying thing, from the marketing standpoint, is that we're simultaneously going to find another group of patients, a potentially larger group with the same disease, who will never take that drug at all. It's going to be a better world, but one in which some business models (cancer therapy!) are going to have to change.


And in a similar vein, reader R. D. writes:

"I have yet to see someone make a rational case for why me-toos are bad. At most, the argument seems to be that if pharma would just stop spending all its time coming up with me-toos, we could get around to curing cancer and parkinsons and stuff. I think that's bunk. You and I both know that any pharma that could come up with cures for things like cancer or parkinsons could start their own mint. The reason they haven't is because it's HARD, not because they prefer to make less money by painting their old pills purple and trying to convince everyone that they're new and improved."


Purple? What on earth can you be talking about? No, the argument he's talking about is one that (in this form) I don't have too much time for, either. The me-too drugs are there to keep the coffers full to pay for the research that doesn't work out, and to tide companies over the dry spells. I can see the objections to the areas where there are six and eight therapies all piled up on top of each other (for example, does the world really need Crestor?) But if Crestor makes money, some of that's going to pay for something new.


And the reason for that touches on another favorite whipping boy: marketing and promotion costs. Keep in mind the inverse relationships between advertising costs, novelty, and the chances of success. A new drug that does something no one's ever seen for a major disease previously thought untreatable - isn't that what makes everyone happy? How much, comparatively, would have to be spent to market such a therapy? There's no competition - it would sell itself! But what are the chances that any of us are going to find and develop such a wonder?


(OK, some of you are saying "Viagra! First on the market, first in the category, promotion out the wazoo!" But keep in mind: no one was sure that men would actually go to their doctor and admit their symptoms - thus the advertising blitz. And Prizer knew, with all the other companies working on PDE subtypes, that competition would be coming soon. They needed all the brand recognition that they could buy.)


Meanwhile, contrast a first-ever wonder drug with, say, the umpteenth statin. It's a crowded field, and you have to spend like crazy to make headway. The thing was a bit lower-risk to develop, since you knew that the rationale was there. But your cost-of-sales figures are going to be uglier, and nothing's ever going to help them.


My point is that a company needs both of these kinds of drugs. You can't hope to live only on the first kind, because they happen so seldom and so unpredictably. And no one's trying to live only on the second kind, either, because you've traded higher costs their for relative security. Everybody developing one of the first class wishes they had some of the second to tide them over. And everyone with drugs in the second class is looking for one from the first.

Comments (0) + TrackBacks (0) | Category: Clinical Trials | Drug Prices | Why Everyone Loves Us

February 26, 2004

Putting A Price on Proving It

Email This Entry

Posted by Derek

I've already had some reader mail (see here) about this article in today's New York Times. It starts out looking like a real pharma-bashing exercise. Up to a point, it is - and up to a point, it's deserved, too. But in the end it's a more subtle piece, not that you'd guess that from the opening paragraphs. (I have my own solution to the problem the article raises, and it will bring joy to no one. Read on.)


The issue is comparability of drugs, especially drugs with the same broad mechanism of action. Look at all the statins or antiinflammatories on the market: is there one that's better than the others? Of course, if you listen to the companies that make them and promote them, the answer is clear. Their product is best! But, as in any other industry, that's not the most reliable guide.


The article uses the example of two marketed forms of the protein erythropoetin, one from Amgen, and one from Johnson and Johnson. J&J's product is about one-third the cost of Amgen's. Is there any reason to pay for the more expensive option? Medicare has asked the National Cancer Institute to run a study to answer that question, but (as the Times points out early and often) there is a provision in the latest Medicare legislation that keeps the program from even using such evidence of functional equivalance in its payment decisions. As you'd imagine, Amgen is arguing that this provision makes the planned Medicare/NCI comparison study a moot point. Why compare?


This would seem like an easy call: the drug companies are slamming the door on something that might cut into profits. Hey, I work here, and I'm sure that that was the motivation, too. But I should add the standard comparisons to other industries at this point, though, and note that car makers are not required to prove that their latest models actually work better than the older ones, or better than the competition's. Nikon doesn't have to run head-to-head trials with Canon, nor Gateway with Dell.


I like those examples, but I realize that there are some other considerations. For one thing, we're talking about public funds here, right? Partly, yes, although the managed-care corporations have a big interest in this, too. I'd add that the government spends a lot of money on goods and services that are not required to be comparison tested (but are selected on the basis of lowest bid.) We'll get back to that topic in a couple of paragraphs. The other big factor is that my car and computer comparisons are discretionary purchases. Health care is treated differently. It's an emotional issue, a life-and-death issue, and it's always going to be held to a different standard than other businesses.


So, let's test! But as the article makes clear, it's not as easy to test these things as you'd think:


. . .Rarely are such studies able to answer all the most important questions. The National Cancer Institute has been mulling the appropriate design for the Aranesp-Procrit trial for nearly two years and will probably need another year before starting the test. . . In the end, more than one trial may be needed, Dr. Feigal (of NCI) said.


Dr. Feigal declined to estimate the cost or size of the eventual trial or trials, but similar tests have cost millions of dollars. Indeed, for comparative trials to be the size needed to measure true differences between drugs, they generally need to be large, lengthy and expensive.


Indeed they do. The article goes on to talk about the hypertension drug comparison study that got such play in the media a few months ago - not least from the New York Times itself. It hasn't settled the question, though. There are still real doubts about which therapy is most effective (for one thing, because patients in the study didn't take more than one type of drug, although in the real world this is a common mode of treatment.) This was a huge study already, and adding arms to assess combination therapies would have bulked it up considerably.


Still, I'm in favor of doing some head-to-head tests, because I think that there are several therapies out there that don't offer much for their price. (I'm looking at you, Nexium!) Here's my proposal - and yes, I'm going to go ahead and treat the drug industry unlike any other. If a company wants to bring out a me-too therapy, it will be required to show evidence of whatever factor differentiates it from the existing agents. The company gets to choose the battlefield: more efficacy? Quicker onset? Fewer follow-up visits to the doctor? Whatever. Pick a reason you're going to promote the drug, and come up with data to back it up. I think we'd end up with fewer me-toos on the market, but we'd lose fewer of them than many critics might think. Many times, drugs that look the same can indeed act differently. Admittedly, it would take some careful clinical work to bring some of the differences out, though.


This change would require a major shift at the FDA. For existing therapeutic modes, you'd need to switch at some point from placebo-controlled trials to competition-controlled trials. Perhaps you could run an initial test-the-water placebo control (after all, these are drugs that have a high chance of working), and from then on you run versus the competition. There are complications - which competitor, for example. But it's possible to do, and it's an idea that has been talked about for a long time.


And who's going to pay for all this? Well, you are (if you're a patient, that is.) Believe me, we're going to pass those costs on, and pronto. Raise the regulatory barrier, pay more money: it's a law of nature. And the lost revenue from the me-too drugs, which have higher chances of success (but still aren't sure things!) will be passed on, too. I think that there are still savings to be realized here - but they're not going to be as big as they seem.

Comments (1) + TrackBacks (0) | Category: Clinical Trials | Drug Prices | Why Everyone Loves Us

February 25, 2004

Ezetimibe, The Press, and More

Email This Entry

Posted by Derek

Credit where it's due! Yesterday I mentioned the original chemist who started the ezetimibe story, but I should note that the drug itself was synthesized by another former colleague of mine, Stuart Rosenblum. He and a host of others developed a huge series of analogs, which built in more acitivity and greater in vivo stability. Just the way drug development is supposed to work, actually.


This drug is also used as an example in a very interesting front-page Wall Street Journal article yesterday. It's a public version of a debate that's been going on inside the industry for a few years now: has the huge increase in compound screening (and compound synthesis) done any good? The article does a pretty good job discussing the issue, although it does mix the two technologies together a bit. It's a very interesting topic, which I'll return to here soon.


And while you're at it, the same issue of the newspaper has (in the Money and Investing section) a nice piece on how drug companies tend to bury news of clinical failures. Different companies handle this differently, of course, but with some of them you really have to watch closely. The article makes the same point I did a while ago - investing in this industry is more of a gamble than most people think. Don't just buy one company's stock if you're looking at biotech and pharma - there's no way you can really know what's going on. Spread your risk.


These articles confirm the Journal's status as the best newspaper when it comes to covering the drug business. The New York Times tries, and sometimes has good work in it, but isn't in the same class. As for magazines, I'd say that Forbes does very well, as does their online site with its copious coverage from Matthew Herper.

Comments (0) + TrackBacks (0) | Category: Cardiovascular Disease | Drug Assays | Drug Industry History

February 24, 2004

The Beginning? It's Right Past the End. . .

Email This Entry

Posted by Derek

There's a paper in the latest issue of Science from a team at Schering-Plough that may have tracked down how the company's cholesterol absorption inhibitor (Zetia, ezetimibe) works. That news really takes me back.


It's been years now, so it won't do any harm to mention that I used to work there. I had a ringside seat for the early years of that project, because it all happened right around the corner from my old lab. Ezetimibe was discovered fortuitously when one of my colleagues synthesized and sent in the original structures of the class for a project targeting a cholesterol handling enzyme known as ACAT. I believe that the in vitro assay was down that week, so the compounds went into the open slots for mouse testing, where they worked better than anything they'd seen. But when the protein assay came back on line, it was discovered that the compounds had no affinity for ACAT at all. Food for thought, that was.


The chemist involved was named Duane Burnett, and a search for "Burnett DA" in Pubmed will send you to most of the chemistry literature on the subject (along with this review). He had indeed hit on some features of a cholesterol binding site (which was his aim, based on blackboard-level structure modeling - no computers involved.) The compounds seemed to hit an unknown target in the small intestine that helped transport dietary cholesterol. The search for the protein involved began in about 1993, and seems to have concluded successfully in 2002-2003, years later than anyone thought it would take.


In the mid-1990s, all the classic methods for tracking down an unknown binding site were tried. The lead structure was biotinylated, modified with radiolabels, photoaffinity tags, and fluorescent groups (along with various combinations of these.) None of these methods identified the target.


They finally tracked down the protein by brute force genomics, using a cDNA library prepared from rat intestinal lining, coupled with sequence searching for the features you'd expect in a transmembrane protein with a steroid binding site. The evidence seems clear that the protein they've found is a key for ezetimibe's actions, but - most oddly - it still doesn't seem to bind to the protein. That would certainly explain the failure of all those modified compounds to pull out the target, but it does make you wonder what's going on. (Is there another real target? But if so, why wasn't that identified through the modified compounds? And so on.)


It took a lot of nerve to go on with that project, and I have to salute the people who kept it going. As with many other successful projects, there were several points along the way where it seemed like the whole effort was going to fail. As it turns out, ezetimibe is one of the main (few?) bright spots in Schering-Plough's portfolio. Merck, their eventual partner for the drug, values it pretty highly, too. I'm glad I got the chance to see it happen.

Credit where it's due! I should note that ezetimibe itself was synthesized by another former colleague of mine, Stuart Rosenblum. He and a host of others developed a huge series of analogs, which built in more acitivity and greater in vivo stability. Just the way drug development is supposed to work, actually.

Comments (0) + TrackBacks (0) | Category: Cardiovascular Disease | Drug Assays | Drug Industry History

February 23, 2004

One Of Us Is Hallucinating

Email This Entry

Posted by Derek

I've been involved in the comments section of another blog, a discussion of pharmaceutical prices which took off from the posts over here. Things have taken a turn that I didn't expect, so I thought I'd run it past the readers of this site. There's a specific misapprehension that I'm running into, one that I've never quite seen before. Am I right to be puzzled? Read on.


Things got underway when the blogger Prometheus 6 picked up a link to my recent drug and research cost pieces from Sebastian Holsclaw's site. (There are a number of comments over there, which are rather easier to deal with than the ones we're coming to.) P6 started off with a blast at the concept of using imputed interest rates to calculate drug costs. We both reference Marginal Revolution on that one, but he's not buying it.


The discussion then split up into a couple of issues. The first was the whole imputed interest / opportunity cost issue, which ended up in another post. Sebastian Holsclaw reappeared to take P6 to task in the comments section, but I don't think that anyone is going to convince anyone else on this issue. This part is worth reading if you want to see some of the difficulties people can have when discussing economics (no doubt my fellow Corantean Arnold Kling gets this kind of thing all the time.)


The other issue arose from P6's final contention in that initial post above. He finished up his take on research costs by saying that "Most of the money spent was federal money." That, as you can imagine, set me off pretty quickly in his comments section. Here's some of what I wrote:


"As for your second question, there are a lot of people claiming that most new drugs are really straight from the NIH, or some such. This is not true. Even if you get an idea for a drug target right out of a paper in the Journal of Biological Chemistry, from an academic lab funded 100% by government grants. . .even then, you're still looking at, on average, that 800 million dollar figure to find a drug and develop it. The amount spent in grant money pales, imputed interest rate and all.


To pick a recent example, the University of Rochester did not develop a COX-2 inhibitor, although they discovered the COX-2 enzyme. Drug companies did - and a majority of companies that tried to make one failed, and all the money they spent to do it is gone.


Academic labs rarely, if ever, come up with a compound that is ready to go to clinical trials. They're doing academic research, as they should: broad fundamental studies that point the direction you should go in to spend your 800 million.


When a drug company does the fundamental research part as well, the costs are even higher."


This spilled over into yet another post, and here's where I've come to realize that I'm not making any headway. P6 seems to believe that most pharmaceuticals should be in the public domain, unless every single bit of every idea along the way was generated inside a drug company. Here's some of his take:


"Let corporations have process patents on the ways they've developed to mass produce the drug, but if it was developed in government funded research the drug itself should be in the public domain."


"My reasoning is, the risk pharmaceutical companies undertake is in creating the processes whereby the drug is produced, packaged and distributed safely (I include testing in this). They are entitled to a process patent. But in most cases the company neither discovered the compound nor ascertained its primary effects."


What's happening here, I thought, is that he's assuming that academic labs produce drugs. And drug companies, it seems, just test them some more and find ways to manufacture them. What a life we'd lead then! So off I went again, in his comments section:


"Unfortunately, you are wrong. In the huge majority of cases (well over 95%, off the top of my head), the compound was discovered by the drug company. Academic labs do not, as a rule, discover drugs. They are not in that business. They discover biological pathways, new behaviors of known proteins, interesting biochemical regulatory mechanisms. Interesting stuff, valuable stuff. But they do not discover drugs. I can think of almost no exceptions. . .


That's how we can patent the compounds, you know. If anyone else has made a compound before and described it in any way, we cannot own the chemical matter by a patent. And we will almost never go ahead with a project unless we're absolutely sure that we own the chemical matter.


Whoever told you otherwise is gravely misinformed. You would do your source a favor by correcting them."


P6's most recent reply to all this begins:


"Sadly, I have no source. I just know what is done in academic research and give it more weight than you."


Well, we can finally agree on something: that's sad, all right. My first impulse was to have him show me some of this academic research that lead right to a new drug, but I wanted to take this over to my readers first, for either irritation or entertainment value as the case may be. As I mentioned, I haven't come across this delusion - that drug companies don't actually discover drugs - at quite this level of virulence before. Is this something widespread? And if it is, how did we in the industry let things get to this state?

Comments (0) + TrackBacks (0) | Category: Drug Prices | Why Everyone Loves Us

February 22, 2004

Reimportation's Just the Beginning

Email This Entry

Posted by Derek

Here are some interesting points raised by reader Duane Oyen, from e-mail and quoted by permission


"I think that any reimportation scheme has to be done on the basis (honest basis, not demagogic political windbaggery, which is where it usually ends up) of free trade- thus causing ME to "get to" YOU- because of the need to eliminate the egregious price discrimination that currently exists. The safety argument, as you state, is a legitimate reason to be cautious, not forbidding. Thus Canada is the perfect guinea pig- even citing NAFTA as the rationale. If a drug company sees that the Canadians are playing price control games with their product, they should be perfectly free to leave that market without having their patents violated, and the US should enforce the IP.


Eventually, the loss of market volume by the company, and the loss of treatment efficacy by the country, will cause the transaction prices to float to a more appropriate level. The ultimate price movements in smaller markets such as Canada will be less significant, as well as relatively slow, giving the lie to those politicians (leftist inheritance limo lib Mark Dayton, bawl your office) who claim that reimportation is a panacea for rising medical costs."


The situation in Canada is: would you rather make a little off your drug up there, or make nothing at all? Companies choose the former - there's no way out. Duane's right that in a more rational situation, the balance between the need for profit and the need for the drug would allow things to reach some sort of equilibrium, but I'm not sure that the warm, windless political conditions needed for that state are ever going to exist.


He goes on to say, in a discussion of what the future of drug pricing holds, that:


"I simply do not see a collapse in drug R&D and discovery. I see relatively significant changes in the business models, but they are already occurring regardless, as more and more of the sclerotic behemoths look to entrepreneurial startups and university spinoffs to feed their pipelines, and drug treatment follows network broadcasting into "narrow-casting". If market changes were established that reduced the ability to internationally price-discriminate, we should do three other things at the same time- a) make adjustments to patent laws (Forbes has made some good suggestions on this over the past couple of years); b) change the NDA process to reduce qualification costs and share risk more fairly (which means that ATLA attorneys have to be reined in a bit); and c) strengthen orphan drug incentives; perhaps even encourage more industry participation in NIH grants."


These are good points. I don't think, though, that there are enough startups and university ideas to keep everyone's pipeline alive - at least, not yet. The in-licensing game is very competitive, and has been for some years now. But he's right that changes in patent law and regulatory approval are basically going to have to take place eventually, and it's time that we starting thinking about what these should be. The idea of strengthening orphan drug-type incentives is floating around in several places, too. This week I'll be talking about a paper that's just appeared that argues for some changes like this in detail.


In coming years, we could be moving to different ways to reward drug discovery other than patent exclusivity and pricing power. It'll be a rough transition, too. The industry had better starting thinking about these issues seriously, or we're going to be dragged, biting and clawing, in directions not of our own choosing.

Comments (0) + TrackBacks (0) | Category: Drug Prices

February 19, 2004

One Of Us Is Hallucinating

Email This Entry

Posted by Derek

I've been involved in the comments section of another blog, a discussion of pharmaceutical prices which took off from the posts over here. Things have taken a turn that I didn't expect, so I thought I'd run it past the readers of this site. There's a specific misapprehension that I'm running into, one that I've never quite seen before. Am I right to be puzzled? Read on.


Things got underway when the blogger Prometheus 6 picked up a link to my recent drug and research cost pieces from Sebastian Holsclaw's site. (There are a number of comments over there, which are rather easier to deal with than the ones we're coming to.) P6 started off with a blast at the concept of using imputed interest rates to calculate drug costs. We both reference Marginal Revolution on that one, but he's not buying it.


The discussion then split up into a couple of issues. The first was the whole imputed interest / opportunity cost issue, which ended up in another post. Sebastian Holsclaw reappeared to take P6 to task in the comments section, but I don't think that anyone is going to convince anyone else on this issue. This part is worth reading if you want to see some of the difficulties people can have when discussing economics (no doubt my fellow Corantean Arnold Kling gets this kind of thing all the time.)


The other issue arose from P6's final contention in that initial post above. He finished up his take on research costs by saying that "Most of the money spent was federal money." That, as you can imagine, set me off pretty quickly in his comments section. Here's some of what I wrote:


"As for your second question, there are a lot of people claiming that most new drugs are really straight from the NIH, or some such. This is not true. Even if you get an idea for a drug target right out of a paper in the Journal of Biological Chemistry, from an academic lab funded 100% by government grants. . .even then, you're still looking at, on average, that 800 million dollar figure to find a drug and develop it. The amount spent in grant money pales, imputed interest rate and all.


To pick a recent example, the University of Rochester did not develop a COX-2 inhibitor, although they discovered the COX-2 enzyme. Drug companies did - and a majority of companies that tried to make one failed, and all the money they spent to do it is gone.


Academic labs rarely, if ever, come up with a compound that is ready to go to clinical trials. They're doing academic research, as they should: broad fundamental studies that point the direction you should go in to spend your 800 million.


When a drug company does the fundamental research part as well, the costs are even higher."


This spilled over into yet another post, and here's where I've come to realize that I'm not making any headway. P6 seems to believe that most pharmaceuticals should be in the public domain, unless every single bit of every idea along the way was generated inside a drug company. Here's some of his take:


"Let corporations have process patents on the ways they've developed to mass produce the drug, but if it was developed in government funded research the drug itself should be in the public domain."


"My reasoning is, the risk pharmaceutical companies undertake is in creating the processes whereby the drug is produced, packaged and distributed safely (I include testing in this). They are entitled to a process patent. But in most cases the company neither discovered the compound nor ascertained its primary effects."


What's happening here, I thought, is that he's assuming that academic labs produce drugs. And drug companies, it seems, just test them some more and find ways to manufacture them. What a life we'd lead then! So off I went again, in his comments section:


"Unfortunately, you are wrong. In the huge majority of cases (well over 95%, off the top of my head), the compound was discovered by the drug company. Academic labs do not, as a rule, discover drugs. They are not in that business. They discover biological pathways, new behaviors of known proteins, interesting biochemical regulatory mechanisms. Interesting stuff, valuable stuff. But they do not discover drugs. I can think of almost no exceptions. . .


That's how we can patent the compounds, you know. If anyone else has made a compound before and described it in any way, we cannot own the chemical matter by a patent. And we will almost never go ahead with a project unless we're absolutely sure that we own the chemical matter.


Whoever told you otherwise is gravely misinformed. You would do your source a favor by correcting them."


P6's most recent reply to all this begins:


"Sadly, I have no source. I just know what is done in academic research and give it more weight than you."


Well, we can finally agree on something: that's sad, all right. My first impulse was to have him show me some of this academic research that lead right to a new drug, but I wanted to take this over to my readers first, for either irritation or entertainment value as the case may be. As I mentioned, I haven't come across this delusion - that drug companies don't actually discover drugs - at quite this level of virulence before. Is this something widespread? And if it is, how did we in the industry let things get to this state?

Comments (0) + TrackBacks (0) | Category:

All Fixed Up

Email This Entry

Posted by Derek

Time to revisit a long-standing topic. You may well have noticed the news items about the FDA's new programs to crack down on prescription drug counterfeiting. That sort of crime is to be expected, given the money to be made - although you really have to wonder about the human status of anyone who turns out fake medicines, full of confectioner's sugar or whatnot. It would be more honest and aboveboard to sneak into someone's hospital room and dig through whatever purses or wallets you could find. Less damaging to the victim, too, for that matter.


So any cost-effective means to cut down on this is welcome. But does a bluff make a sound when it's being called? Regular readers will know that I've spoken out several times about drug reimportation from Canada, and how I oppose it on economic grounds. But my industry's trade association, PhRMA, has taken their stand on the drug safety part of the issue. "Don't import those contaminated drugs!" goes the cry. (How many of the counterfeits seized to date actually came through Canadian pharmacies, I wonder?)


As I've said, one problem with this line of attack is that it doesn't get at the real heart of the issue, which is the cost of research and the money that needs to be made to fund it. (And yes, I know, the costs of marketing, too - but keep in mind that companies, in whatever industry, don't spend that kind of marketing money year after year without getting a good return on it as well. We can debate how much advertising drug companies should be allowed to do, but don't assume that we're throwing the money out the window.) There will be more next week on drug costs and research spending (the mail keeps on coming!), but no matter what, I think we can assume that the two are somehow related.


No, the real problem with the drug-safety argument is that it makes reimportation seem like an otherwise harmless thing, that just suffers from this unfortunate side problem. What happens if someone actually fixes it? What's my industry going to be able to say then? There is a backup plan, right, guys?

Comments (0) + TrackBacks (0) | Category:

February 18, 2004

How Drugs Die

Email This Entry

Posted by Derek

Everyone in the industry would like to do something about the failure rate of drugs in clinical trials. It would be far better to have not spent the time and money on these candidates, and the regret just increases as you move further down the process. A Phase I failure is painful; a Phase III failure can affect the future of the whole company.


So why do these drugs fall out? Hugo Kubinyi, in last August's Nature Revews Drug Discovery suggests that it's not for the reasons that we think. As he notes, there are two widely cited studies that have suggested that a good 40% of clinical failures are due to poor pharmacokinetics. That area is also known in the trade as ADME, for Absorption, Distribution, Metabolism, and Excretion, for the four things that happen to a drug once it's dosed. And we have an awful time predicting all four of them.


Of the four, we have the best handle on metabolism. In the preclinical phase, we expose compounds to preparations from human liver cells, and that gives a useful guide to what's going to happen to them in man. We also expose advanced compounds to human liver tissue itself, which isn't exactly a standard item of commerce, but serves as a more exacting test. Most of the time, these (along with animal studies) keep us from too many surprises about how a compound is going to be broken down. But the other three categories are very close to being black boxes. Dosing in dogs is considered the best model for oral dosing in humans for these, but there are still surprises all the time.


That 40% figure has inspired a lot of hand-wringing, and a lot of expenditure. But Kubinyi says that it's probably wrong. Going back over the data sets, he says that the sample set is skewed by the inclusion of an inappapropriately large group of anti-infective compounds with poor properties. If you adjust to a real-world proportion, you get an ADME failure rate of only 7%.


Now, when this paper came out, I think that there was consternation all over the drug industry. (There sure was among some of my co-workers.) The ADME problem has been common knowledge for years now, it was disturbing to think that it wasn't even there. So disturbing, it seems, that many people have just decided to ignore Kubinyi's contention and carry on as if nothing had happened. There have been big investments in ways to model and predict these properties, and I think that many of these programs have a momentum of their own, which might not be slowed down by mere facts.


The natural question is what Kubinyi thinks might be our real problem. In his adjusted data set, 46% of all failures result from lack of efficacy in Phase II. He admits that some of these (in either approach to the data) might still reflect bad pharmacokinetics, but still maintains that poor PK has made a much smaller contribution than everyone believes. Here's his drug development failure breakdown, which makes his point:


46% drop out from lack of efficacy
17% from animal toxicity (beyond the usual preclinical tox)
16% from adverse events in humans
7% from bad ADME properties
7% from commercial decisions
7% from other miscellaneous reasons

Comments (0) + TrackBacks (0) | Category: Drug Assays | Drug Development | Pharmacokinetics | Toxicology

February 16, 2004

The Old Days, Some Older Than Others

Email This Entry

Posted by Derek

Last Wednesday's Wall Street Journal had a very interesting article on the front page: "In Two Generations, Drug Research Sees a Big Shift." It profiles Leo Sternbach, discoverer of the diazepams (Valium being the most famous) and other drugs, and his son, Dan Sternbach of GlaxoSmithKline. The elder Sternbach has been at this stuff since 1940, so he's seen the lot. (The article isn't online, but an earlier account of his research is here.)


He began work in the era of open buckets, wooden paddles, and trying drugs out on yourself to see if they did anything. After one of those, he recounts, "For two days I was not at all well." In my first job, I overlapped with just a few people from the last of the era, and they had similar stories. The thought of eating one of my own research compounds has always given me the shakes. No thanks - not before a whole long list of mice, rats, and larger mammals have had their turn. Believe me, no one does that today, and anyone who did would be asking for huge amounts of trouble. It's interesting to note, though, that the practice doesn't seem to have done Sternbach any long-term harm, since he's 95 and all. I should note, for balance's sake, that other chemists from even earlier days are known to have poisoned themselves but good through such techniques.


I found the article particularly interesting because Dan Sternbach was a professor of mine when I was in graduate school at Duke in the mid-1980s. He left for greener Glaxo pastures while I was there, but I did take his class in frontier molecular orbital theory. (That's yet another of the things I learned in grad school whose applications since then I can count on my fingers. It's hard to explain briefly for my non-chemist readers, but here's a PDF to give you the flavor of it.) I would have found it pretty interesting to know, back in 1984, that Sternbach and I would end up at competing drug companies. At least once, to my knowledge, we've been competing in the same area of research at the same time. His years at Duke meant that he started in industry just three years before I did, actually, although with his running start he certainly came in at a higher level.


One Sternbach story I have is when a group of us were grading exam papers for a section of sophomore organic chemistry he was teaching. We attacked it as a team, each person taking on a set of questions and grading the whole class's answers to them, passing the papers on to the next person for them to check their section. One of us hit a paper where the student had made it about halfway through an answer, and had scribbled in a frustrated "I can't remember the goddamn equation!" We all enjoyed that one, but as his paper made it around the table, it turned out that he had a couple more similar fits: "Don't know the goddamn reaction!" and so on. When it came time to total up the scores, Sternbach looked at the guy's exam and mused "I should just write on there: '148 out of 200, Goddamnit!' "

Comments (0) + TrackBacks (0) | Category: Drug Industry History | Press Coverage

February 15, 2004

Drug Prices and Costs - From the Mail

Email This Entry

Posted by Derek

Here are some of the responses to the recent posts on drug prices and research costs. First off, from Dave S.:


"I'm a skeptic on the idea that the proceeds from the temporary legal monopoly represented by a patent is, in fact, used by pharmaceutical companies to finance R&D. Yes, I know it stands to reason and it's the prevailing wisdom, but is it true? One way to investigate this would be to see how the companies themselves view R&D. This, in turn, may be determined by how real R&D expenditures vary with real revenue. If, as real revenue rises, R&D expenditures rise either proportionally or faster, it would be a reasonable conclusion that patents do support R&D. If, however, R&D expenditures grow more slowly than revenues or not at all, it would be equally reasonable to conclude that the prevailing wisdom is wrong and that patents do not in fact support R&D but just raise profits, which nice as it is for the companies involved is not the constitutional purpose of a patent."


Interesting point. I'll see if I can round up some figures, but I'm not sure how good they'll be. Like any other industry, pharma companies do all tricks with earnings statements that they possibly can, in order to tell Wall St. a coherent story. And there's also the tendency to think, as the money comes in, that "Hey, we don't need to increase research that much. Look how well we're doing with what we have!" Conversely, on the way back down, as revenues fall, R&D isn't always the first thing to get cut. By that time, it's clear that the only thing that will save the situation is for some new profitable breakthrough to emerge from the labs. But let's see what the numbers say.


There have been several letters along the lines of this one, from Mark S.:


"What I find baffling is that no one is pointing out who the enemies really are. Americans are subsidizing the world. Subsidizing AIDS drugs in Africa may be defensible, but subsidizing cholesterol drugs in Europe isn't. I'd like to see some information about what would happen to American drug costs if Europe (and Canada) were just cut off. Obviously some money is getting into pharma coffers, the question is how much? What sort of profits are European Pharmas making in America vs Europe (i.e. are they extorting Americans, too)?"


Well, the larger European companies depend on revenue from American sales, too, I can assure you of that. As for cutting off the European market, it would be a difficult, although intruiging, thing to calculate. I don't know of any company that publicly breaks down cost-of-sales for the two markets, though (although I'm sure that these figures are worked out internally.) Marks' suggestion:


"While not a big fan of regulation, I think it would be fair to regulate drug prices in America this way: The cost to Americans is the lowest cost negotiated with any other country. It doesn't cap drug costs, it just evens them out. Drugs could cost anything, but everyone would be paying the same amount for the same product. Pharmas are free not to sell to any country if doing so would lower the US price. They are also free to negotiate different prices for different countries, the US would always match the lowest."


I think that this would fall victim to a sort of Gresham's Law - bad prices would drive out the good. Countries with national health plans already can use their muscle to negotiate prices that couldn't be sustained across the worldwide market - "You want to make a tiny amount of money here, or nothing at all?" As long as the US runs a different sort of market, things wouldn't mesh too well. Here's an approach to that problem, from reader Clark H:


"If the US, the last of the free markets, ever starts allowing reimports or does "National Bargaining", I would suggest that the Drug Companies start the bargaining BEFORE they have finished their R&D. When they still have leverage. The problem with any product with a large up-front cost and relatively small operating costs is that it is at a huge disadvantage bargaining with a small number of customers after all the costs are "sunk". The customer knows the hand of the supplier. The supplier can't walk away because any positive cash flow is a good thing at that point. The drug companies should get to phase 2, when they know roughly what to expect (after all they are designing the phase 3 endpoints), and then bargain. If the countries aren't willing to pay an adequate rate, they walk away having spent only the cost that would otherwise have been spent. Not a perfect solution, but. . ."


I think that there's something to this, although the Phase III failures would throw a wrench into the works. Of course, that's what they do now, actually. I wonder, though, if what would happen under this plan is that countries would just apply a standard calculation for what they thought the Phase III trials and regulatory approvals would cost, which would take things right back to the "already knowing the hand of the supplier" stage.


Well, the mail on this topic doesn't show much sign of letting up, so we'll keep it open and revisit it in a few days. Brainstorms appreciated.

Comments (0) + TrackBacks (0) | Category: Drug Prices

Easterbrook Post Updated

Email This Entry

Posted by Derek

I wanted to point out the update at the end of the "Cold Equations" post below, in case anyone's interested. There's been a lot of mail on this one, as you'd imagine - but so far, nothing defending Gregg Easterbrook's analysis. If anyone out there is minded to do that, I'd like to hear it. Nor have I heard from the man himself, and his original post remains in the same inaccurate state in which it was born. Is it a blog if you never update a post or revisit a topic?

Comments (0) + TrackBacks (0) | Category:

February 12, 2004

More on Prices, High and Otherwise

Email This Entry

Posted by Derek

The key questions raised in the e-mail I quoted in the last post are: is it fair to fund drug research through high prices on drugs? And especially, is it fair to do so by raising prices on individual drugs, rather than across the board?


My answer to the first question, as you'd guess, is "yeah boy!" I have some disagreements with the assumptions that Nick H. is making. For one thing, the money spent on research isn't necessarily an "investment cost" in the way that some might think of it. Now, overall, it's true that we expect a return on research money, so you could classify the whole think as an investment that way. But research (unlike, say, bonds) involves a lot of sunk costs. Once spent, they can't be recovered or converted back to cash. The investment equivalent is a stock that goes to zero. A drug company's "investment portfolio" consists of a bunch of Webvans and the occasional Microsoft.


But that analogy isn't quite right, either, because it gives too much credit to the individual drugs. Microsoft can, by its own actions, try to preserve its profits indefinitely (and boy, don't they ever.) But every drug is a wasting asset, essentially. The key difference is intellectual property. A drug's patent will always expire after a few years, although companies will try every contortion they can think of to delay the day. The longer the R&D&A (A for approval!) process takes, the shorter the lifetime left on the patent. It is very common, and has been for some years now, that a drug chews up more than half of its patent life before it ever makes it to market. So every drug is born onto the market from a deep pit of sunk costs, and has a relatively short time to make them up (and pay everyone's salaries, and fund the current research projects, some of which had better pay off to keep the whole thing going.)


On to the second question. I think that Nick's main objection is that prices get raised on individual drugs, rather than across the board. His point can be illustrated by Abbott's price hike on Norvir, a drug that has (from their perspective) been underperforming. If Abbott needs more money, why should only the HIV patients pay up? (I know, they've assured Medicaid, etc., that the price hike doesn't apply to them - but some HIV patients or their health care providers are going to pay more, though. Otherwise, why raise the price? Unless it's just to make Norvir more expensive for the competition, which is one possibility that can't be shaken off.)


One answer - not the best one- would be to turn the question around: why should someone who takes one of Abbott's other drugs have to pay for the problems they've having with Norvir? This argument doesn't hold up too well, though - after all, we pass on the research costs of the whole drug portfolio to everyone. Why not pass on the sales shortfalls?


Well, in a larger sense, we do. But I'd say that companies - and not just drug companies - raise prices where they think that they can, when they think they can. A company may decide that it's feasible to raise prices in one market, and not another. To use Nick's supermarket analogy from the previous post, should we require all the products in a supermarket to have the same profit margin? I would actually expect a supermarket to raise its profits by increasing prices on the items it thought would be most likely to bring in the most increased revenue. (That probably wouldn't be the bread and flour, though.)


This is what the marketing types call "pricing power," and it varies a lot. A drug company has quite a bit of it, during that time its drug are patent protected. Countering that, of course, is the negotiating clout of Medicare, HMOs, and other bulk purchasers. That's the general tug-of-war of the market (making some allowances for Medicare.) Really countering a company's pricing power, outside the US, is the governmental price-setting ability of national health plans. That's where the market mechanism is breaking down, to my eyes. Many people see intellectual property laws as a thumb on the scale in favor of the drug companies. But at least patent protection goes away.


(There's a lot of interesting e-mail coming in on this subject, and I'll bring up some of the points people raise in another post. I should mention that Nick, the person who starting this topic off, has a blog of his own, too.)

Comments (0) + TrackBacks (0) | Category: Drug Prices

The Cold Equations

Email This Entry

Posted by Derek

Here's a question for the folks at The New Republic: are you really sure that Gregg Easterbrook's blog is a good idea? Just checking. He's already stepped into so many steaming mudholes that I'm starting to think that he's one of those born-to-be-edited types. (Or would that help?)


I say this in response to today's entry, a diatribe against Eli Lilly over the suicide of one of their Phase I clinical volunteers. A 19-year-old student at a nearby Bible college with no known history of depression, she was one of one hundred normal subjects who were taking Lilly's new antidepressant, Cymbalta (duloxetine). She had completed the dosing protocol, which started early in January, going up in dosage and gradually back down to placebo. At the time of her suicide, she had been on placebo for four days.


Easterbrooks's upset about this, and on the face of it, who can blame him? The suicide of a young, apparently healthy person is disturbing to anyone. But he's even more upset at Eli Lilly:

"TRACI JOHNSON, 1985-2004: I've never read anything so heartless as the claim by the Eli Lilly drug company that its experimental anti-depressant Cymbalta has absolutely nothing to do with the fact that a healthy, apparently non-suicidal 19-year-old woman committed suicide while living in an Eli Lilly dormitory and serving as a subject in a trial of Cymbalta. No wait, I have read something still more heartless--Lilly's claim, to The New York Times, that although 5 of the 9,000 people who have taken Cymbalta, whose chemical name is duloxetine, have killed themselves, the company sees no connection. "We have not been able to discern any signal between duloxetine and suicide," an Eli Lilly spokesman told the Times."


This piece is the victim of a common fallacy - that is, taking what's in the New York Times as a mirror of reality. Let me recommend, in this case, the Indianapolis Star, which is at least out there on the ground in Indiana. There one might read that Michael Turek of Lilly attended chapel at the bible college on Sunday morning and spoke to the student body, quite possibly with enough heart to satisfy Gregg Easterbrook.


I just hate to add myself to the roll call of the heartless, but perhaps this Eli Lilly spokesman is actually telling the truth. Easterbrook isn't buying, though. Second paragraph, with my emphasis added:


Can this be anything other than a straight corporate lie? The suicide rate for all Americans is 10.8 per 100,000, according to the Surgeon General. This means 5 suicides would be expected in a group of 46,000 people, not in a group of 9,000. In a group of 9,000, one person would be expected to commit suicide, which suggests Lilly's own trials show that Cymbalta is causing subjects to kill themselves at five times the expected rate. The rates are suicides per 100,000 per year, but presumably Cymbalta trial subjects have not taken the drug for longer than a year. And some in the trials group may have symptoms of depression, which could put them at higher risk for suicide than the population as a whole. Still, five suicides in a group of 9,000 is way too high to be coincidence."


Bullshit. Some "may have had symptoms of depression"? Gregg, those other four deaths were all in phase II trials. They were in depressed patients. You can't compare those figures to the ones derived from the general population. And the suicide rate of those depressed patients taking Cymbalta is lower than those who are taking placebo - as it had damn well better be. It's actually ten times lower than the suicide rate of those taking existing antidepressants. The Indianapolis Star could have told you that.


It's true that such drugs may work differently in normal patients versus depressed ones, and it's also true that the effects may be different in younger patients (the subject of a lot of controversy in the field at the moment.) I can't deny any of that, and thus can't rule out that such an effect may have caused Traci Johnson's suicide. But neither can I rule it in. On the basis of the clinical data Lilly has, there's no way to know, and there's no reason to assume it. Their spokesman is correct.


Let's go back to Easterbrook for a wrap-up:


"Some people, including some young people, really do need anti-depressants, and society benefits from pharmaceutical companies researching such drugs. But pharmaceutical research isn't just a formality to win marketing approval--when trials show a drug has dangerous side effects, the research is supposed to stop. . .


Dr. Alan Breier, Lilly's chief medical officer, went so far as to make small of the young woman's death, telling the Times: "Most people who commit suicide in the general population leave people asking these kinds of questions. And just because this happens while someone is taking a drug doesn't mean the drug caused it."


What a malicious, cold-blooded thing for a corporation to say. Lilly has essentially just told the parents of a 19-year-old woman that it doesn't care that their daughter is gone. Physicians should fall to the floor and weep if they ask a completely healthy young person to take an experimental drug, and a month later she is dead. . ."


Dr. Breier is, I'm sorry to report, absolutely correct. Lilly has been told to stop enrolling patients in their latest studies on Cymbalta while the FDA studies this latest news. That's the right decision, too - but I think that they'll find that there's no way to know if there's a connection or not. Post hoc ergo propter hoc is the name of the fallacy that would tell you that we already know, but that's exactly what Breier is trying to say.


And he's not, as far as I can see, being malicious at all. And if he's being cold-blooded, well, then so am I. And so should everyone in clinical research. Does Cymbalta cause younger patients to become suicidal? We don't know, and we will never know until a lot of young people have taken it. I hate to put it that starkly, but it's true. You have to be willing to find the answers to questions like these, or you can't help anyone.


I am sorry that Traci Johnson is dead. But if we fall to the floor and weep, as Easterbrook recommends, we're neglecting our duty to the living. He's right - when trials show that a drug has dangerous side effects, the research is supposed to stop. But - excuse the cold-bloodedness - Traci Johnson's suicide does not establish this. Would we be walking away from a drug that could wind up preventing suicides instead of causing them?


"One owes respect to the living - to the dead one owes only the truth." - Voltaire.


Update: Check out this blog post, which begins: "I've been on anti-depressants for over two years now, so I feel qualified to offer some additional comments."


And I missed several Easterbrook/blogosphere dustups in my list above. Try this one, this one (also here), and this one (the January 13th post.)

Comments (0) + TrackBacks (0) | Category:

February 11, 2004

The Contact Sport of Cost Accounting

Email This Entry

Posted by Derek

I have just enough time to post an interesting e-mail, from Nick H. in the Netherlands, responding to the links I posted the other day about research costs.


"I agree that accountants calculate costs in the way described, but accountants do a lot of accounting in ways which aren't always quite fair to all stakeholders. I have to disagree that all research costs should necessarily be paid fully by the buyer(s) of a limited number of products. The money spent during research/on research could also be seen as an investment cost. It is quite normal to defray part of investments by increasing the price of product X, it is just as normal to defray part of those costs in other ways (e.g. lower dividends, accepting lower profits or by raising the costs of all products to a lower extent). Saying this is the only (fair/possible) way to do things just isn't cutting it.


. . .the real argument which we can find buried way back in the mists of time is why a limited number of customers should pay for all costs incurred for no other reason than to increase business returns as well as increasing those returns. If your supermarket puts up a new building you'd be well miffed if they raised the price of ONLY the bread and flour to defray the building costs. Especially if the raised it to the extent needed to pay back all of the building costs within 10 years. It would be seen as unfair and wrong by pretty much everyone anywhere that this supermarket would hold just this limited (local) part of the population at ransom by suddenly raising the price of an essential condiment by a large margin in order to pay for what ultimately is mainly to the benefit of the supermarket itself. I hope you agree that there is a grain of reason in the madness of suggesting that costs should be divided more evenly over the various stakeholders."


I don't agree with all his points, but I thought I'd put them out there to stimulate debate. I'll post my response to these tomorrow. The phrases "intellectual property" and "sunk costs" will appear prominently.


(I should mention that my post-a-day schedule will probably be hitting a few potholes over the next few weeks. I'll keep the content flowing (I like doing it!), but there will be weeks when only three or four new posts appear.)

Comments (0) + TrackBacks (0) | Category: Drug Prices

February 9, 2004

I will do such things - What they are yet I know not. . .

Email This Entry

Posted by Derek

Igor Landau of Aventis must have been seen a performance of King Lear recently, to judge from his tactics to fend off Sanofi's hostile takeover bid. Or since the following also has a "Do you feel lucky?" ring to it, maybe they've been watching old Clint Eastwood movies instead. Surely those have been dubbed into French. From a Dow Jones article by Hollister Hovey:


Because Sanofi's stock-and-cash bid was hostile, Sanofi doesn't have access to the fine print of Aventis' marketing deals. Many of Aventis' products that it sells or is developing through joint ventures or co-marketing deals have so- called change of ownership clauses written into the contracts, said Mr. Landau. If the Aventis partners which have those clauses in their deals decide they don't like Sanofi, that may give them an option to regain rights on the products, noted Chief Operating Officer Richard Markham. Aventis won't say which products they are. . .but Mr. Markham said the products are "material enough" to mention the risk.


Meanwhile, Sanofi is running newspaper ads with pictures of sick children, implying that they won't get well unless the merger goes through. Reminds me of the old National Lampoon "Buy this magazine or we'll shoot this dog" cover. I have to say that I'm enjoying watching all this. Doubtless it wouldn't be quite as entertaining if my job were one of the lightweight objects being blown around in the whirlwind.


I have no useful predictions of whether this deal will go through. For one thing, companies make idiotic acquisitions all the time, terrible potlatches of money and ego. Sometimes these things just take on a life of their own - folks get so wrapped up in trying to win that they forget why they were playing the game. The encouraging words from the investment banking community don't help much with that problem. It's like asking a contractor whether he thinks you really need to add those three rooms to your house.


I have the opposite bias: I think that pharmaceutical mergers have an unencouraging history. Even the ones that have more or less worked out have done so after a terrible period of unfocused disruption, during which many good ideas hit the wall and many good people hit the exits. As I've said before, the whole basis for big drug mergers seems specious to me. I refuse to believe that research productivity scales in anything close to a linear way. After a company gets to a certain size, I think it's more likely to fit an exponential curve - but in the wrong inflection, mind you.


Even if Sanofi does succeed, it probably won't be on the terms they wanted. The gap between Sanofi's offer and Aventis's actual market value is widening again. Here's a recap of how those two have compared since the bid was announced. (Note that in this chart, the original bid was at -3.6%, a premium to Aventis's stock price at the time.) The shares of both companies have gone in opposite directions since then. That represents a lot of betting that Sanofi will need to come up with a better offer, and that someone else may well do the same. Here's hoping that everyone involved keeps their wits. But if there were a stock-market proxy for that proposition, I'd go short on it.

Comments (0) + TrackBacks (0) | Category: Business and Markets

February 8, 2004

Darn Those R&D Costs, Anyway

Email This Entry

Posted by Derek

I wanted to pass along a couple of recent articles that address drug pricing and research costs. It's a subject that attracts nonsense like a cloud of gnats; every so often you have to shoo them off. Here's a fine post from Alex Tabarrok at Marginal Revolution:

In 2003, Joseph DiMasi, Ronald Hansen, and Henry Grabowski published an important paper in the highly-regarded Journal of Health Economics that estimated that the average cost of developing a new drug was around $805 million dollars. Hal Pawluk at Blog Critics repeats some nonsense from Public Citizen to claim that high research costs for pharmaceuticals are a myth and that this paper in particular is part of a conspiracy of pharmaceutical companies to raise prices. Frankly, the comments of the critics are laughable but not everyone sees the joke so I will explain. . .


And his explanation is a good one, right on target. I can speak from experience that this is exactly the way that costs are calculated in the industry, and quite rightly. Check it out and see what you think.

I can also strongly recommend this column from the Wall Street Journal's Holman Jenkins. It's a broadside against the dime-store populism that's turning up in Presdential campaign speeches. Here's a sample:

If this column sounds like one four years ago, that's because Democrats are running against their usual list of "enemy" industries. The party's standard trope is that you're being denied things you need and deserve because enemies are keeping them from you, cheap drugs being today's case in point.


Let's make sense of the industry once more for a Democratic presidential cadre now reaching a high pitch of populist dudgeon. There's a reason analysts, investors and pharmaceutical reps talk about a "pipeline." In one end goes a bunch of money, and out comes a dribble of products years later. The metaphor is also useful in understanding drug pricing. Whatever comes out the end, whether it's nose drops or a chemotherapy drug, is priced at whatever level will allow its maximum contribution to recouping all the money that went into the front end of the pipe.


Just so. It's one of those darn businesses, which, As Is Well Known, are invariably run as a conspiracy against the little guy. (The huge population that actually works for these operations are, by definition, not little guys, it seems. Not after they issue you your Monopoly-man top hat along with your employee ID card.)

Comments (1) + TrackBacks (0) | Category: Drug Prices

February 5, 2004

Sic Transit Gloria Mundi

Email This Entry

Posted by Derek

So Entremed is finally giving up on its celebrated peptide drug candidates, angiostatin and endostatin. I'm sorry to see what the company, and its employees have been through, but I'm even sorrier when I think about what cancer patients have been through during this story. Especially those of them who read that (in)famous New York Times story back in 1998, with the (in)famous James Watson quote about curing cancer in two years. (He might as well wear a sign: Loosest Cannon in the Room - Come Here For Quotes!)


I wrote about this whole issue back on my Lagniappe site a couple of years ago, which led to an interesting exchange with Mickey Kaus about whether the Times article had been vindicated or not. Kaus is a wonderful political journalist, but would be the first to admit that he's no pharmaceutical expert, and his take was that the focus on angiogenesis made the Times article look better, in retrospect. My contention was that angiogenesis was already known to be a good area to work in. This was thanks in good part, it's true, to Judah Folkman, who was also a main subject of the article. But the emphasis on angiostatin and endostatin was perverse.


By the time the Times ran their front-page above-the-fold horn-honker, legions of people had already decided that Folkman was on to something. And they were working on better compounds to try the theory out. The problem with Folkman's peptides was that they were, well, peptides, and unusually painful ones to work with, at that. Angiostatin and endostatin are not that easy to produce, to purify, to store, or to dose, as opposed to the small molecules that were also already in clinical trials.


Once you dose peptides, they get hammered, chain-sawed, and burned down to the ground. I love coming up with verbs to describe the process, I have to admit. If you saw a plot of blood levels after dosing a typical peptide drug, you'd see just what I mean. It takes a pretty specialized protein to circulate around after an i.v. dose without getting enzymatically ripped to shreds, and as for oral dosing, you can forget it. The gut is very efficient at grinding every protein it sees down to its component amino acids.


So Entremed's compounds were flying into a stiff breeze from the beginning. And, as so often happens, anti-angiogenesis as a cancer therapy has turned out to be more complicated than anyone thought. (If you work in the drug industry, you could take that last phrase and turn it into a keyboard macro to save time. It's practically our motto.) There are many, many possible targets in the field, and we're not that sure which ones are most likely to work. And a lot of them are so biologically similar that it's hard to make selective drugs that work against them, anyway, thus many clinical candidates turn out to have a large and ill-defined footprint. To add to the confusion, different types of tumors express different amounts of these various target proteins, and their relative importance is surely all over the place as well. Not that we know yet, of course. And I'm only talking about the cases where vascularization is thought to be important; many other sorts of tumors are poor candidates for anti-angiogeneic therapies right from the start.


No, these compounds have been doomed for years, as far as I can see. They were long shots even when the Times sent Entremed on its insane ride up to nearly $100/share. I remember scoffing at the article when it came out, and I remember trying to go short the company's stock (couldn't get the shares to borrow.) But while all this was happening, people were getting their hopes up, desperate patients who thought that this might be their chance for survival. How many of them who read that article are still alive today?


I'm normally a pretty optimistic person, but I make an exception for my work. We should never let people get their hopes up. Not until we're really, really sure. We don't know enough. This whole situation wasn't Entremed's fault, and it certainly wasn't Judah Folkman's. This one gets chalked up to the New York Times.

Comments (0) + TrackBacks (0) | Category: Cardiovascular Disease | Drug Development | Press Coverage

February 4, 2004

Our Friend, the Carbon-Aluminum Bond

Email This Entry

Posted by Derek

"Twenty-five years of being a laboratory chemist, says Gregory Hlatky today, and this is the first time I've had an incident." Hey, maybe he's been doing the wrong kind of chemistry. Some kinds can almost guarantee you an incident every month or two!


Actually, the kind of chemistry he does (organometallics) is already pretty lively, and I have to say that I'm impressed by his 25-year safety record in a field like that. But I'm not surprised that it was an organoaluminum compound that took off on him, because I've had several of them do the same thing to me (without injury, fortunately.)


And the most nerve-wracking part of them was the time delay. Most reactive compounds are very forthright about their reactivity. They burst into flame on exposure to air (like tertiary-butyllithium, or for the hard-core pyromaniacs, the dialkyl zincs.) Or they give off great clouds of choking fumes (I can recommend neat titanium tetrachloride for those who want to experience this special effect - the one molar solution in dichloromethane is for dilettantes,) or hiss and splatter violently if they encounter water (chlorosulfonic acid is a winner in that category.) At any rate, you know very quickly, if you didn't already, what kind of substance you're dealing with.


But the alkylaluminum compounds have their coy ways. I recall a large aluminum alkyne reaction that I set up in graduate school, one of the once-and-for-all reactions which get scaled up a little more than is prudent. I don't believe that I've told this story on the blog, so this one can go into the file with my other lab stories.


With this reaction, the fun started early. I first had to add a large amount of n-butyllithium, which is less reliably pyrotechnic than the tertiary kind, and I'd done that by running the solution into a Pyrex dropping funnel. That, for the non-chemists in the readership, is one of these. As the stuff dripped slowly into the reaction, the BuLi had dried into a crust on the glass tip of the funnel. At the end of the addition, I had to pour in a larger quantity of toluene and switch the funnel to another piece of glassware, so I just grabbed the thing and swapped it out. Whereupon it burst into lovely orange/purple flames.


Well, they went out shortly. But I was standing there, pouring my (flammable) toluene with one hand, and holding this flaming funnel with the other, thinking that this would be a good time for a member of the department safety committee to show up. After that, the next steps of the reaction went along relatively quietly, and eventually it was time to quench the reaction. I did that by very carefully adding a few drops of methanol to the liter or so of solution. Nothing. So I added a few more.


Nothing. I waited to see if anything would happen. Nothing did. A brief squirt of methanol this time. Zilch. I was starting to wonder if there was going to be any reaction at all - surely there was some leftover organoaluminum stuff that needed to be quenched. A longer, more vigorous squirt of methanol. No sign of life.


Then, at the very bottom of the flask, down near the magnetic stir bar, a bubble formed and rose to the surface. And another. Several more. A stream, several streams, a vigorous fizzing mass that came roaring and foaming up the sides of the flask - Well, I managed to catch most of it in a bucket. When the bubbling started to really roll along, I had bolted for something to catch things in, because it was clear that the reaction could be on its way to a spectacular conclusion. It didn't disappoint me.


Like most once-and-for-all reactions, I had to do that one again, eventually. But the next time, at least, I was ready for it.

Comments (0) + TrackBacks (0) | Category: How Not to Do It

February 3, 2004

Ricin Redux

Email This Entry

Posted by Derek

Now that the suspected ricin in the Senate (and White House?) has been confirmed, I thought I'd repost a version of something I wrote about a year ago on my previous site, Lagniappe. (This was written after British authorities had rounded up several suspects in London who had some ricin of their own.) So what is the stuff, and what kind of threat is it?

Ricin's a protein from castor beans - yep, the same ones used to prepare castor oil. The parent plant is sometimes used as a warm-weather ornamental, and used to be an industrial crop. The leaves aren't a problem, but the beans contain up to 5% ricin, which is a rather high yield for a natural product. It's quite toxic, although there are certainly worse things out there. Botulinum toxin, for example, is a thousand times more potent, but you can't grow anerobic bacteria very well in your back yard (and they're not very ornamental, either.)

The purification methods for ricin are in the open literature, and aren't particularly challenging. I'm not going to go over them, though, in the interest of not making its isolation any easier than it already is - it's already probably one of the easiest toxins to isolate. For that matter, you can order various forms of it from biochemical supply houses. It's quite cheap, by the standards of protein natural products (which are usually priced rather steeply.)

And what does the stuff do? Briefly, it's a very potent inhibitor of protein synthesis, which it accomplishes by attacking one subunit of the ribosome (the central RNA-to-protein machinery of the cell.) Rather than just binding to ribosomes and gumming them up, ricin is actually an enzyme all by itself. It tears up a specific adenine base in the ribosomal RNA, which disables the whole thing, and then it moves on to the next ribosome. One ricin molecule can turn that reaction over thousands of times, and needless to say, a cell can't lose thousands of ribosomes and expect to survive.

Ricin's a reasonably large protein, and as a weapon it suffers from the defects of large proteins. The least dangerous way to be exposed to it is by eating it, since most of it gets digested, and much of the rest has trouble crossing from the gut into the bloodstream. In rodents, the worst way to be exposed is by inhalation. Oral dosing is about 4000 times less potent. The assumption is that if ricin were weaponized, it would be treated like anthrax spores and dispersed for maximum effect. The US and Britain carried out research that led to a prototype of a ricin bomb during World War II, just another one of many nasty weapons that actually didn't get used in that conflict. (It's hard to imagine the second World War being even worse than it was, but it had the potential.)

Needless to say, there's not a whole lot of public data on just how toxic ricin might be in that aerosol form, and its effects would certainly depend on particle size, static charge, and all the other variables we learned about during the anthrax scare. We sp have a single public data point about injected ricin, though: Georgi Markov, a Bulgarian exile who worked for Radio Free Europe. One day in 1978, he felt a sharp pain as a stranger poked him with the tip of an umbrella. He began to feel ill within a few hours, and three days later, he was dead. A small pellet containing ricin had been injected into the muscle of his leg, as it turns out, in one of the more exotic assassinations known to have been carried out by the KGB. The best guess is that about a half a milligram proved lethal.

Well, that sounds pretty bad - but consider that terrorists are unlikely to be able to give masses of people intramuscular injections. And if they want to use inhalation, which is certainly the way to cause real damage with the stuff, they're faced with manufacturing problems similar to the use of anthrax spores. It wouldn't be easy, fortunately.

Ricin's not particularly water-soluble, so dumping it into a reservoir would be a waste of time. And adulterating food would be almost useless. It takes a good handful of the beans themselves to kill an adult (and they have to be crunched up, too, because whole beans tend to pass, well, pretty much unchanged through the digestive tract.) A back-of-the-envelope calculation for the pure toxin suggests that it would take a gram or two to reliably kill someone by ingestion. That adds up to a few hundred casualties per pound of ricin, but only if you can get all your victims to eat enough of it. (And I've no idea how stable the protein is in hot food, but I'd have to think that it would be inactivated pretty quickly.)

The key to this latest ricin incident will be whether the stuff had been processed to the level of the anthrax attack material. If so, we have a potentially serious problem, but so far I don't see any sign that this is the case. Otherwise, this could well be the work of some disgruntled and/or deranged amateur. Those types I think we can deal with.

Comments (0) + TrackBacks (0) | Category: Chem/Bio Warfare

February 2, 2004

How to Be an Inventor

Email This Entry

Posted by Derek

While I'm talking about inventorship on patents, I should note that there's a factor that doesn't get the attention it deserves: luck. Well, not the public attention, anyway. But talk to any group of researchers about who gets on which patent, or whose lab produced the most active compounds in any given project, and the word will inevitably come up.>/p>

There's something to that talk. We really don't know how to make active drugs on demand. Otherwise, I can assure you, we'd be making a heck of a lot more of them than we are. There are too many variables, and too many things we just don't understand. In that sense, every drug that's made it to market has been the beneficiary of plenty of luck indeed. At the research stage, you never know who's going to produce the anointed clinical candidate.


And as for who gets on the patent, well, if we don't know what compounds are going to be good ones when the project starts, then it's hard to say what the final patent claims are going to look like, either. People divide up the structure and work on different areas, some of which are going to take off and most of which aren't. When you begin work on a project, you really have no idea where the winds are going to take you.


I'm making it sound like inventorship is all a matter of chance, but that's only partly true. To be listed on a drug patent, you have to have made a distinct non-obvious contribution to the claimed invention, which in our case is usually a new series of compounds. Some idea of yours has to be one of the things that's being claimed. A good mental test is to imagine a lawyer asking you what your part of the invention is. If you can't immediately point to something specific, you're very likely not an inventor.


With that in mind, you can see which way things are headed as a project comes closer to maturity, and judge for yourself if you're likely to be included on the patent(s). Researchers who are alert to this sort of thing often adjust their work accordingly, to have something to point to come patent time. This means that it's harder, especially in some organizations, for lab associates to be listed as inventors. If someone is serving as merely the proverbial pair of hands, they're not going to be on there.


And they shouldn't be. Even if you make the wonder drug that saves the whole project, you're not an inventor if you're just doing what someone else thought up and told you to do. I've always tried to make sure that people who report to me understand the legal requirements of inventorship, and I give them room to think of (and work on) their own ideas. You get better research done that way, too, I might add.


An extra patent note - Gregory Hlatky responds to yesterday's post by pointing out that although we US researchers don't get a cut of the profits from our patents, we don't have to pay the fees associated with filing them, either. He's right that these can be quite substantial, epecially for the ones that you're filing all over the world. I take his point. But I regard those fees as a cost of doing business, like buying reagents (which I'm glad I'm not paying for out of my own pocket, either.) Of course, in the end, we employees are (quite properly) paying for all of these, since that money, derived from business profits, could otherwise be used to increase everyone's salary, among other things. The same argument applies, of course, to the money that would be spent giving inventors a cut of said profits, but I think it could be money well spent.

Comments (0) + TrackBacks (0) | Category: Patents and IP

February 1, 2004

Ricin Redux

Email This Entry

Posted by Derek

Now that the suspected ricin in the Senate (and White House?) has been confirmed, I thought I'd repost a version of something I wrote about a year ago on my previous site, Lagniappe. (This was written after British authorities had rounded up several suspects in London who had some ricin of their own.) So what is the stuff, and what kind of threat is it?


Ricin's a protein from castor beans - yep, the same ones used to prepare castor oil. The parent plant is sometimes used as a warm-weather ornamental, and used to be an industrial crop. The leaves aren't a problem, but the beans contain up to 5% ricin, which is a rather high yield for a natural product. It's quite toxic, although there are certainly worse things out there. Botulinum toxin, for example, is a thousand times more potent, but you can't grow anerobic bacteria very well in your back yard (and they're not very ornamental, either.)


The purification methods for ricin are in the open literature, and aren't particularly challenging. I'm not going to go over them, though, in the interest of not making its isolation any easier than it already is - it's already probably one of the easiest toxins to isolate. For that matter, you can order various forms of it from biochemical supply houses. It's quite cheap, by the standards of protein natural products (which are usually priced rather steeply.)


And what does the stuff do? Briefly, it's a very potent inhibitor of protein synthesis, which it accomplishes by attacking one subunit of the ribosome (the central RNA-to-protein machinery of the cell.) Rather than just binding to ribosomes and gumming them up, ricin is actually an enzyme all by itself. It tears up a specific adenine base in the ribosomal RNA, which disables the whole thing, and then it moves on to the next ribosome. One ricin molecule can turn that reaction over thousands of times, and needless to say, a cell can't lose thousands of ribosomes and expect to survive.


Ricin's a reasonably large protein, and as a weapon it suffers from the defects of large proteins. The least dangerous way to be exposed to it is by eating it, since most of it gets digested, and much of the rest has trouble crossing from the gut into the bloodstream. In rodents, the worst way to be exposed is by inhalation. Oral dosing is about 4000 times less potent. The assumption is that if ricin were weaponized, it would be treated like anthrax spores and dispersed for maximum effect. The US and Britain carried out research that led to a prototype of a ricin bomb during World War II, just another one of many nasty weapons that actually didn't get used in that conflict. (It's hard to imagine the second World War being even worse than it was, but it had the potential.)


Needless to say, there's not a whole lot of public data on just how toxic ricin might be in that aerosol form, and its effects would certainly depend on particle size, static charge, and all the other variables we learned about during the anthrax scare. We sp have a single public data point about injected ricin, though: Georgi Markov, a Bulgarian exile who worked for Radio Free Europe. One day in 1978, he felt a sharp pain as a stranger poked him with the tip of an umbrella. He began to feel ill within a few hours, and three days later, he was dead. A small pellet containing ricin had been injected into the muscle of his leg, as it turns out, in one of the more exotic assassinations known to have been carried out by the KGB. The best guess is that about a half a milligram proved lethal.


Well, that sounds pretty bad - but consider that terrorists are unlikely to be able to give masses of people intramuscular injections. And if they want to use inhalation, which is certainly the way to cause real damage with the stuff, they're faced with manufacturing problems similar to the use of anthrax spores. It wouldn't be easy, fortunately.


Ricin's not particularly water-soluble, so dumping it into a reservoir would be a waste of time. And adulterating food would be almost useless. It takes a good handful of the beans themselves to kill an adult (and they have to be crunched up, too, because whole beans tend to pass, well, pretty much unchanged through the digestive tract.) A back-of-the-envelope calculation for the pure toxin suggests that it would take a gram or two to reliably kill someone by ingestion. That adds up to a few hundred casualties per pound of ricin, but only if you can get all your victims to eat enough of it. (And I've no idea how stable the protein is in hot food, but I'd have to think that it would be inactivated pretty quickly._


The key to this latest ricin incident will be whether the stuff had been processed to the level of the anthrax attack material. If so, we have a potentially serious problem, but so far I don't see any sign that this is the case. Otherwise, this could well be the work of some disgruntled and/or deranged amateur. Those types I think we can deal with.

Comments (0) + TrackBacks (0) | Category:

Deal Me In

Email This Entry

Posted by Derek

The inventor of the blue-light diode, Shuji Nakamura, has been fighting for some years now for a bigger piece of the profits from his invention. Since his original payment was about US $189, that's easy to imagine. The New York Times article on the case is fine, as far as it goes, but it contains one very puzzling paragraph:


On patent filings, Japanese companies list their names above the inventor's; Western companies typically file patents on behalf of their employees and pay them a percentage of the profits.


My readers around the drug industry are all going "Since @#$!*! when?" I can tell you that all you get at a large pharma company is a hearty handshake. Oh, if your drug does make it to market, it's certainly good for your career, true. But a percentage of the profits? Not in the US. But there are places.


Take Germany, for example. Under German law, inventions produced by employees of corporations are still their property, which is transferred to the company in exchange for specific payment. There are all sorts of situations that arise from this (the link above will take you through them) but the result in the drug industry is that the inventors on a patent get a share of the profits from the marketed drug. Interestingly, a similar law has been proposed in the UK, and is now being debated.


This sets up some major struggles about inventorship, as you'd imagine. I'm told there have been some real brawls over the years. In a US drug company, being an inventor on a patent is quite desirable. But in Germany, it can change your life.


As someone who's been an inventor on many issued patents, I can tell you that I'd certainly support such a system in the US. There's a real case to be made for it, and not just by folks who could profit (hey, none of my patents has ever made a dime for anyone.) Over on The Bottom Line, fellow Corantean Arnold Kling has commented on the patent idea that was discussed in the post below (which he'd also noticed, just before I did.) He proposes a greater use of prizes as an incentive to innovation, which I think is a fine idea. A piece of the action would be a good start.

Comments (0) + TrackBacks (0) | Category: