|
|
About this site
|
|
Here we'll review recent developments in drug discovery and medicine and the IP issues and financial implications they have, along with general thoughts about research. Also likely to make an appearance: occasional digressions into useful topics like which lab reagents smell the worst.
|
|
|
About this author
|
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.
|
|
|
|
Cipro in Iraq?
News reports have surfaced that some Iraqi soldiers have been found to be carrying the antibiotic Cipro as part of their kit. This has immediately led to speculation that there is some connection with the possible use of anthrax.
Unfortunately, that can't be ruled out. Iraq most certainly seems to have produced a nightmarish amount of anthrax during its biological weapons program, and that's just the stuff we know about. Anthrax can take a few days to come on, and thus isn't as useful in a fast-evolving battlefield situation (which this one sure is.) But you can't really be sure that it wouldn't be used if it were around.
However, the presence of Cipro isn't a direct link to anthrax. It's true that the drug got a lot of publicity in the fall and winter of 2001-02 as an anthrax therapy, but it's not the only effective antibiotic against the bacterium. Just about everything in the fluoroquinolone class of drugs should be effective, and plenty of others as well. If it was found with other supplies, it's possible that this was just part of someone's idea of a good medical kit to have (to guard against gangrene, etc.) But if the antibiotic was there all by itself, then one does have to wonder. . .worth watching. |
How Do You ID a Chemical Weapons Plant?
As I write, there are reports all over the news wires that a suspected chemical weapons facility has been found in Iraq. If this holds up - and there's no certainly that it will, just as there's no certainly about any early reports in any war - how can the presence of these weapons be confirmed?
Well, if there's still some in the pipeline, that's the obvious shortcut. It's not a trivial matter to just open the stopcock and see what comes out, but you've if you've got the material there, it's impossible to mistake it once you've analyzed it. Chemical weapons are small molecules, which are easily identified - an undergraduate could tell some of them from their spectral data.
Residues inside the equipment are also pretty diagnostic. It's impossible to get all the stuff out of any chemical production facility without a thorough and meticulous cleaning. Unless great care has been taken, you're always going to have traces of what was being produced. (And even if such cleaning has taken place, there could well be washes and waste piled up - it has to go somewhere unless there's a very good incineration facility on site.)
And, finally, there's feedstock. No chemical plant operates without supplies of its raw starting materials around. If you find tanks of thiodiglycol, for example, their owners are going to have some explaining to do. That gets used in ballpoint pen ink, sure, but if it's in a well-hidden site, you're looking at mustard gas precursor. Meanwhile, the nerve agents all involve phosphorus compounds which can only be plausibly confused with insecticide starting materials. And the further down the synthesis they are, the further the structures diverge from something that only kills flies.
No, establishing that a given facility is involved in making chemical weapons isn't going to be hard at all. The tough part is finding them, and taking them. |
Antisense Ungood
Antisense company Isis Pharmaceuticals had a bad day yesterday. Well, actually, they've probably had several bad days recently, it's just that yesterday the rest of us got to hear about it. The stock had the bad day today. They've been plugging away for years at antisense technology, where a complementary "backwards" strand of nucleic acid is intoduced to gum up the expression of a specific gene. The idea is a slick one, and has been a slick one for a long time now. It's just that the execution is such a bear.
DNA and RNA molecules themselves have a pretty short half-life out in the body. They're not the most stable things in the world, not when they're ripped away from all the proteins that help them out in the cell nucleus, and the body probably isn't real keen on having naked nucleic acids wandering around all over the place, either. A major challenge in antisense has been getting reasonably long-lasting analogs that can still penetrate the cell membrane from outside (and get through the nuclear membrane after that.)
This latest compound was a joint venture with Eli Lilly on an antisense to protein kinase C-alpha, PKC-alpha to the trade. This is a reasonable cancer target, and it's one that small molecule therapies have had very little luck attacking. (PKC subtypes, which are legion, have been a real tar pit for drug development in several disease areas.) They made it all the way to Phase III, definitely the most advance thing that Isis has.
And then the wheels fell off. Phase III trials are larger than Phase IIs, and tend to be done under more real-world conditions. This studied non-small-cell lung cancer patients, half of whom got standard chemotherapy and half of whom got that plus the Isis drug, Affinitak. And there was no benefit whatsover. Of course, there have been some half-hearted attempts at putting a good face on the data, but from what I've seen, Affinitak is a goner - at least against NSCLC.
Truth to tell, this wasn't unexpected. Isis and Lilly were using a pretty strict standard (patient survival, not some secondary marker) against a very difficult cancer type. And word was getting out back in December that the trial might not look good.
All this is still a disappointment, but don't count Isis out yet. Affinitak was from an earlier generation of antisense, and they now have better structural classes that seem to have better activity. These are in earlier trials, so their main challenge is to hold on until those come in. But it's been tough: Isis had a good-sized deal with Merck for a diabetes therapy (another enzyme that no one has a good small molecule against.) Many people thought that one could pay off, but Merck dropped it back in December (which must have been a bad month for these guys.) They're still in there pitching, though, with deals with Amgen and Pfizer.
Someday antisense will pay off. . .maybe. The big question is whether RNA interference will come from behind to take over the same therapeutic space. Is that one going to be easier, or is its fate to be just as tough as this stuff is? |
France and the European Court
No, despite what the last couple of days have been like around here, this isn't going to be about Iraq (at least, not directly.) A co-worker pointed me to this article at Genomeweb. It's about the EU's Biotech Directive, which is an attempt to get the various member countries in line on matters of biotech policy and intellectual property law.
That's no small order, because many EU member countries still have their own patent offices. There's a central European Patent Office, but applications still issue as separate patents in Germany, England, and so on. There's usually not much difference between them, since they generally come from an EPO application, but there are areas of disagreement - especially on patenability. ANd when a patent gets challenged, it can get fought out country by country. Sometimes the EPO steps in and clears things up eventually, but sometimes not.
Among other things, the Biotech Directive says that gene sequences are patentable, as long as a clear utility is demonstrated for them. That's essentially the position of the US PTO, at least after they clarified things a couple of years back. The UK and others have implemented this, and it's now part of patent law in those countries. But several EU countries haven't gotten around to it yet. Including France. . .and they've just passed their own biotech law, which blatantly contradicts the EU directive at several key points. For example, gene sequences are specifically declared non-patentable.
According to the article, the French aren't even keeping up a pretense:
But the French government has made no attempt to suggest that the two pieces of legislation are compatible. In fact, French Health Minister Professor Matthei, a sponsor of the bill, suggests that the best way forward would be for the Biotech Directive to be interpreted in accordance with the (French) Bioethics Bill.
That's a good one. Wouldn't you really rather do it the French way, now that you've thought about it? All you other countries, who negotiated and implemented the binding EU-wide policy - didn't you really mean to do it the way France thought it should be done?
There's room for disagreement on gene sequence patentability, and plenty of subtle points to keep everyone occupied. But this isn't subtle at all. It's, well, it's. . .unilateral. That's the word I was looking for. In theory, France should now be dragged before the European Court. Won't that be a spectacle. . .
|
Chemical Attack
As we draw closer to the resolution of the Iraq situation, it's time for me to make good on my promise to update the chemical warfare posts that I did last fall (which are here, starting on the post dated Wednesday, Sept. 11. The date was not completely a coincidence.)
I think it quite possible that a military campaign in Iraq would cause Saddam Hussein's forces to use unconventional weapons. It's not a good decision, objectively, but the Iraqi government has made similarly poor decisions in the past. If things come to that, here are the possible chemical agents: 1. Hydrogen cyanide: industrial chemical, although not produced and stored in large quantities as such. Can be defended against by standard protective equipment.
Effectiveness: minimal against reasonable protection. Wasn't even useful in World War I.
Likelihood of use: If Iraq has weaponized HCN, they're more stupid or desperate than anyone's realized. Despite its reputation, HCN is a very poor choice for a war gas. Delivering enough of it rapidly enough to be effective is a problem not worth solving.
2. Chlorine: industrial gas, available in huge quantities. Made by the ton by anyone with a chlor-alkali chemical capacity, which can be quite low-tech. Powerful smell, not something that sneaks up on anyone. Can be defended against by standard protective equipment.
Effectiveness: minimal against reasonable protection. Considered obsolete as a weapon.
Likelihood of use: Iraq has plenty of chlorine. But I strongly doubt that it would be used in battle. If they're going to go chemical against a force like the one arrayed against them now, they're going to go all the way.
3. Phosgene: industrial gas, available in large quantities. Can be made by the ton by anyone with a supply of chlorine. Odd smell, though not as immediately detectable as chlorine. Can be defended against by standard protective equipment. Effectiveness: minimal against reasonable protection. Considered obsolete as a weapon.
Likelihood of use: Iraq may have phosgene. But I also doubt that it would be used, for the same reason as given for chlorine above.
4. Mustard Gas: not really a gas, but an oily persistant liquid. It stays around long enough to be a problem for everyone involved, including the side that initially uses it. It's cheap to make, and requires only what a standard chemical industry would have already. It keeps indefinitely if reasonably clean. But it can be defended against by standard protective equipment.
Effectiveness: minimal, despite its reputation, at least against well-equipped soldiers. It can take a long time to incapacitate or kill, making it more of a terror weapon. As such, it's potentially horrific if used on civilians.
Likelihood of use: Iraq most certainly has produced this, and it was used against Iran. Its use now would be an extremely poor decision, because it would accomplish nothing except an even more overwhelming response. But it can't be ruled out.
5. Tabun (GA): one of the more volatile nerve gases, and the first one to be produced on a large scale (by a hideous process; see my September discussion for more.)
Effectiveness: well, all the nerve agents are pretty damned effective. That's the problem. But good protective gear will negate them, as long as it's coupled with discipline and good training.
Likelihood of use: Tabun is the "training wheel" nerve gas, and is considered to be essentially obsolete. Even if a country or group has never produced a nerve agent before, they'd be more likely to come up with Sarin. Saddam Hussein has worse.
5. Sarin (GB): still used as a nerve agent. Like the others, it's very stable on storage if it's clean to start with. This was what Aum released into the Tokyo subway, although their material wasn't as clean as it could have been, fortunately.
Effectiveness: worse than Tabun. Good protective gear will negate it.
Likelihood of use: Sarin has certainly been produced in Iraq, and could see use if it's still around. I'd say that VX is the most likely thing in this department.
5. Soman (GD): a little-used nerve agent, although it was produced on large scale and stockpiled during the Cold War. Effectiveness: worse than Tabun and Sarin, overall. Good protective gear will negate it.
Likelihood of use: Unsure if this has ever been made in Iraq. Worse has, though.
5. VX: the bottom of the barrel, or the top, depending on the state of your soul. This was the state of the art before the 1971 agreement, and there's been no confirmed public report of anything past it.
Effectiveness: less volatile than the other nerve agents above, and more of a problem by skin contact than by inhalation. Skin contact is more than enough. Likelihood of use: Appears to have been used in the infamous attack on the Kurds. If Iraq is going to deploy a nerve agent, which I fervantly hope they are not, then this is the one they're going to use.
So, how would they use any of these? The standard method has been by artillery, but there's a limited amount of agent that can be delivered that way. They can be dropped by aircraft, as Iraq has horribly shown by its own experience. The drone that's caused all the weapons-inspection furor would be one method, but it would have to be low- (and slow-)flying to disperse a chemical agent. (Biologicals could be a different story.) But there aren't going to be any Iraqi aircraft getting off the ground, or not for very damned long. (That's not bragging, nor is it wishful thinking, just a statement of fact.)
Chemical attack would be an act of revenge or desperation on the part of Iraq, and (after the first shock and suprise) would have no effect on the outcome of the conflict. The problem is, we can't rule revenge or desperation out. The effects would be worst on any civilian population in the area, of course, which is yet another reason to hope that we don't seem these weapons used.
For more on these, if you can deal with more, see this large PDF. An excellent one-stop reference.
|
Vaxgen and Its Numbers
The March 7 issue of Science has a summary of Vaxgen's presentation of their HIV vaccine results, under the discouraging headline of "Vaccine Results Lose Significance Under Scrutiny." The contents of the article will hold few surprises for readers of the posts here from the last week of February. The pitfalls of subgroup analysis are discussed, sometimes in exactly the same terms as my contributors and I used:
Cornell University's John Moor, a longtime critic of the vaccine and an expert on HIV antibodies, finds this reasoning absurd. "Lumping Blacks and Asians together is biological rubbish," says Moore. "They might as well do a subgroup analysis based on signs of the zodiac."
An interesting wrinkle occurred during that week after the criticism started. Many pointed out that the P values (used as a measure of statistical significance) really should have been corrected for how many subgroups were split out. A conservative way to do this, the Bonferroni correction, is to just multiply the P values by the number of subgroups. That way you have to have much more significance in the underlying data before you decide that you're seeing something real.
Did Vaxgen do this? Their initial press release sure didn't make it clear, but it didn't look as if they had. Then on February 27, their CEO told an investor conference that sure, you bet, they'd done that. In fact, he said, "a conservative version was applied, and it had no impact on statistical significance." This is the sound of someone who doesn't appear to understand what he's talking about. Of course doing the correction had an impact - if you're lucky, it doesn't push your results into nonsignificance, but there has to be an impact.
Later that day, Vaxgen issued a a press release "in response to media inquiries." As they said, seemingly through gritted teeth:
"The number of required adjustments for VaxGen's subgroup analyses is subject to interpretation and there are a variety of methods to calculate those adjustments. The company cannot predict the impact these adjustments may have on the findings since that determination will ultimately rest with regulatory authorities."
They're presenting at a Keystone meeting in Banff in about two weeks. Should be a mighty interesting session...
|
I Claim Stuff That's Good For What Ails You
The U. of Rochester patent defeat that I wrote about on Friday has some pretty interesting implications for Ariad's attempt to sew up the NF-kB transcriptional pathway. Much as I'd like for Rochester to have lost on general principles, the general principle that you can patent medical methods of treatment (and entire biological pathways) remains intact. It'll probably take statute law to change that. The reason Rochester lost was, basically, "lack of enablement."
That, translated from patent-speak, means that they didn't show how to do what they were claiming, and showed no evidence that they'd done it (or even knew how to.) It's one thing to claim the COX-2 enzyme; it's quite another to claim any compound that works by inhibiting it. Rochester had the enzyme in hand, and they'd done the background work to show what it could be used for. But they didn't have anything to accomplish those uses. As the judge's opinion said, they basically were expressing a wish, or outlining a plan - not carrying it out.
And you have to do that, or at least show enough work that you can claim that you know how to. Just saying that you think it would be a good idea isn't enough. Of course, the Rochester patent had already been issued in this form, which just goes to show you what the patent office is like these days. Now, if they'd shown some examples of compounds that inhibited COX-2, they'd have had a much stronger case - and the more compounds, and the more data they had on their effects, the better.
The people most likely to have plenty of data like that are, well, drug companies. It's much harder to invalidate big pharma patents, because they're generally backed up with more enablement. Not impossible, not at all - just harder. So will Ariad's hold up? Based on this decision, I'd be worried if I were them. They have no compounds in their patent at all, at least none that I recall seeing. So how can they say that all their claims of medical treatment are enabled? The whole patent looked to me like a gigantic shopping list.
And this brings up another point. Back in December, there was an article in The New Republic from two former editors of the New England Journal of Medicine. And it just raked the drug industry over the cheese grater, back and forth for pages on end. I kept meaning to take this article apart, but never got around to it (for one thing, it goes on at a beastly, New Yorker-ish length.) But there's one point that the authors harped on which always gets my blood up - the accusation that drug companies steal all their research from publicly funded discoveries.
I'd be the last to deny that we get a lot of useful stuff from NIH-funded research projects (and many others.) If someone working off an NIH grant discoverers that knocking out the XYZase gene leads to doubled lifespan with minty fresh breath, you can bet that we'll all try to make an inhibitor of its protein product (and try to make a fortune off it.) But that's not piracy - that's science. We build off each other's discoveries.
The attitude seems to be that it's only the people who find the pathway that have discovered anything. Someone who comes up with a drug that uses said pathway is just freeloading. On the contrary: it ain't that easy. If it were, I can assure you that we'd have a lot more useful drugs out there (and we'd all be making a lot more money.) Going from a biological target to a drug is about as hard a job as you could ask for. As I've mentioned before, I've been doing this since 1989, and I've never worked on anything that's come close to ever hitting the market. That's very typical. Years go by between working on something that even gets into a clinical trial, much less one that makes it all the way.
My message to anything who thinks differently is: come on down and try it. Get yourself some venture capital and grab some of that easy money. We're not scooping much of it up right now, for some reason - it's all here waiting for you.
|
Couldn't Resist
My post on Scifinder and modern literature searching the other day reminds me of a story. Back in about 1985 or so, I became the grad student in my group who searched CAS (Chemical Abstracts Service) Online. We got a 90% discount, but it was still considered something that the boss wasn't going to let everyone go wild with.
As I mentioned, trying this out for the first time, I felt like I'd suddenly been given voodoo powers. Searching through bound volumes was slower, of course, and less thorough - but I quickly became aware that there were online searches that weren't just slower than the corresponding offline ones, they were downright impossible that way. (Multiple variable atoms and chain lengths, for example, would have been just nightmarish with the hard copies.)
Well, one day I was doing a search for one of the new grad students in the group. I ran the query and got a reasonable list of hits. As I scrolled down the screen, she suddenly pointed at the title of one of them: "That's it! That's exactly what I need!" I thought quickly - pointing at some of the arcane codes that littered each citation, I said "But that's going to be a hard one to get. . ."
"What! How? Why? I've got to have that one!" "Well, do you see that part where it says "JPC"? That means that it's from the Journal of Pakistani Chemistry." (Sheer bull, of course, but I thought it rather inspired.) "Journal of. . .what?" "And that part that says "UR?" That means that the paper is written in Urdu."
Well, I let her rant for a bit, then broke the welcome news that the paper was actually from something like the Journal of Physical Chemistry - actually, I think I'd rather read that non-existent Pakistani journal instead, but at least we had J-Fizz-Chem right over on the shelf. And they don't publish in Urdu, although most folks would need about three issues to notice the switch even if they made it mandatory. So she calmed right down, after switching from berating Chemical Abstracts to berating me.
And a side effect of that one was that she wouldn't trust a word out of my mouth for about four months. "What time is it? " "Noon." "Yeah, sure, you expect me to believe that?"
|
Ker-thunk. Ker-thunk. Ker-thunk.
Schering-Plough was out with yet another revised earnings forecast. I've lost count of how many times they've done this in the last few months, but it's easy to remember what direction they've all been. (I have some SGP stock; believe me, it's easy to remember. It's a rule, by the way, that every time at least one financial news source has to run the news under the head "Schering Gets Ploughed.")
The problem has been that sales of Claritin haven't been what the company had expected. That's just weirdly incompetent (the Wall St. Journal called their situation "a textbook case of bad transition planning.") I think that they should have made provisions for Claritin revenue to drop off, say, 80% in the first three months after it went off patent. That would have been a blow to the stock at the time, but a preferable one to rolling down the entire flight of stairs one by one, which is what's been happening.
After all, look at what's happened that last few years with big patent expirations. Sales just vanish; managed care sees to that, and quite rightly. That's what an efficient market does. It responds quickly to price signals, turning quite literally on a dime. How could Schering-Plough think that things would be any different in their case?
(Well, OK, I know how: the usual way. A bunch of people got together and repeated the sales forecasts to each other until everyone believed it. Companies do all kinds of crazy stuff that way. But wasn't there anyone around who could burst the bubble?)
|
What's A Patent Buy You?
There have been some very large sums transferred over the answer to that question. Yesterday's decision in the long-running dispute over the University of Rochester's COX-2 patent may be ready to join the list. That's the enzyme behind money-spinners like Vioxx (from Merck) and Celebrex (now Pfizer's.)
It's beyond doubt that the university team discovered the enzyme, and realized its possible use in anti-inflammatory treatment. (Anyone who didn't catch on to that part wouldn't have been competent enough to have found the enzyme in the first place, but that's neither here nor there.) They filed a broad patent, and broad claims issued - including language that would seem to have claimed treatment of COX-2 mediated diseases with whatever inhibitor anyone might discover.
Long-time readers will recall that these patent claims really drive me insane. (See my articles on June 26 and July 2nd about Ariad's patent abuses, eventually to be followed up on.) Until a few years ago, no one tried to claim this sort of thing, but now that the USPTO has let these things issue, everyone's in on the act. Universities, companies, random strangers: they're all patenting their enzymes, the sequence that codes for them, cells that express them and vectors to make 'em do so, every assay that might use any of the above, the use of any of those assays to discover any drug, and the treatment of any disease with any of said drugs. That's only slightly exaggerated. Very, very slightly.
The Rochester group didn't have any chemical matter that might have been a COX-2 inhibitor, and they had no idea of what one might look like or how to make it. But they went after the companies that did the work of finding and developing them, wanting a generous piece of the pie. A summary judgement has tossed that on the trash heap - well, for another few days or weeks, until they appeal. Who knows when this will eventually get resolved?
Matthew Herper at Forbes does a fine job, by the way, of pointing out that Pfizer has been doing something very similar to its competitors who are working on drugs that go through the same pathway as Viagra. He hits it on the head:
Essentially, Pfizer draws the line of "invention" at the moment when researchers actually figure out how to make a drug. The University of Rochester picked a different "eureka" moment, when scientists figured out exactly what the drug should do. The question is whether the law should allow either of these discoveries to be protected as inventions.
Both of these patents are weird, or at least counterintuitive. It would be easy to make a host of philosophical or legalistic arguments for either patent. But the purpose of patents is not really all that complicated. They are supposed to encourage invention by giving companies a degree of monopoly over their products, but not squash it by preventing anyone from doing anything similar ever.
For this reason, Pfizer's PDE-5 patent is a bad idea. Having a whole host of medicines to treat a disorder is a good idea, even if they are similar. . .
Preach it! I couldn't agree more. We're going to ruin our industry with this nonsense if we don't watch out.
|
Life of a Salesman
Unfortunately, I'm even more pressed for time tonight than I was last night. I have to present details of a project tomorrow to various higher-ups, which is always a bit of a tightrope act. After all, every drug research project has potentially fatal problems. Really, all of them do. It's just a question of whether any of these are going to grow to the size that they'll be trouble. That odd behavior at the high dose. . .that time the scale-up of the key reaction didn't work. . .that time the formulations people had trouble dissolving a new batch of compound. . .the odd low stability number that showed up in a metabolism test. . .any of these could turn into a project-breaker. More likely, they're just odd bits of noise or minor irritants, but you're never really sure until the end.
So presenting a project is an exercise in what to leave out. You want to give an accurate impression of things (well, most people do; there are exceptions.) After all, if the whole thing does go down in flames in another couple of months, you don't want bewildered managers ringing your phone because they thought everything was fine. But you also don't want to drag in every possible complication, because that'll make it look like resources should be pulled away now before any more time gets wasted.
I try to aim for "promising, but with some problems to overcome." That way, if things work out, you and your people look good for having made it through. And if the project melts down, which so many of them do, then you've warned people that it was a real possibility. A corollary to that worldview is to never, ever take on a project that looks like it can't fail. They all can fail. And when one of those can't-miss things misses, they don't ask what was wrong with the project; they ask what was wrong with you.
|
Old Paper
Little time to blog this evening - I've got some work to do here at Lagniappe Control, and I'm going to work early tomorrow to prepare some spiffy-looking overheads. At least, they're supposed to be. This is another time when it would be convenient to be able to talk about my work, because it's touched recently on several very interesting points. Unfortunately, a long discussion of those would be unwise, and there's no way I can strip out the proprietary aspects and leave much that's worthwhile.
I spent a fair amount of time today using the same tool that Bob Standaert surely used to track down those structures in the post below - SciFinder, the desktop application from the folks at Chemical Abstracts. For someone like me, who remembers searching the 10-year CA index volumes by hand back in the early 1980s, a program like SciFinder still has some magic to it. You really had to be there, I guess: somewhere off in the distance, a radio plays Thomas Dolby or "Come On, Eileen" as,on the other side of a set of double doors, a younger version of me in a "Nuke the Whales" T-shirt pages through the Formula Index looking for whatever name CA was using in the 8th Decennial Index for, let's see, C-eleven H-fifteen N. . .hmm, they named it as a derivative of acetonitrile, who'da thought - pretty soon they're going to have everything named as a derivative of methane, then the alphabetical index will be mostly a bunch of volumes from M to M. . .how about the Ninth? What? Nobody made this stuff for ten years? Can't be a good sign. . .
. . .and then I'd be reaching for a pen to write down a bunch of volume and citation numbers, using those to pull down all the individual abstract volumes and look through them. I used to do that all the time, even when I first started in industry - all chemists did. But it's been years now since I even picked up a bound volume of Chemical Abstracts, and who knows when (or if) I ever will again. I'd miss 'em more if using them hadn't been such a pain. I'm more than happy to give up the thrill of the chase, believe me - there are plenty of more useful chases to occupy my time.
Couple something like SciFinder to a bunch of online journal subscriptions, and there's an insane amount of information ready to coming hosing right to your desk. The American Chemical Society now has every back issue of all its journals available electronically, which means that I won't be picking up many old bound volumes of the Journal of Medicinal Chemistry any time soon. And it also means that I'll need to find something to do with all those late-80s and early-90s issues of the Journal of Organic Chemistry that are cluttering up a cabinet in my office in piles of dark royal-blue covers. Any takers?
|
Ready For These?
Robert Staendart from the University of Illinois at Chicago writes that the graphics from the Ready.gov site include some chemical structures, and they're interesting choices. On the page that link takes you to is a compound that would probably be made from 4-piperidinone, an intermediate that's an old friend of mine (and to many other medicinal chemists.) This structure is missing a hydrogen off one of the nitrogens - if you put it back where it should be and search the databases, Standaert says, you find that:
The N-H compound is known but obscure; it appears in the registry as the free base ([56587-63-4)], a citrate ([56587-64-5]), and as a mixture with droperidol (an antipsychotic), but there are no references to any of these since '67.
The structure shows up more clearly in this PDF. It certainly looks like a CNS-active compound - if I had a nickel for every piperidine structure that shows up in the CNS literature I'd have - well, one heck of a lot of nickels. In fact, as Standaert points out, this one is very close to the recently-notorious fentanyl. It looks like they took off a phenyl group for aesthetics and just left the nitrogen dangling without a substituent.
Meanwhile, on the biological warfare part of the site, the PDF file available has an unusual structure in the background, too. It's drawn oddly, with the carbonyl bonds moved down into the structure. But as Standaert points out in his e-mail, it's surely a structural fragment common to the aflatoxins.
Now, that makes a bit more sense as a bioweapon than fentanyl does as a chemical one. Aflatoxins can be pretty nasty, with acute toxicity as well as long-term liver damage and carcinogenic effects.They're not much good in a battlefield situation, since they're pretty slow-acting (years, in the liver mechanism.) But as a terror weapon, they'd unfortunately fill the bill.
You'd have to make the stuff yourself by fermentation, which fortunately isn't trivial at all. Synthesis is out of the question; you'd need a team of people working for months to make milligrams of the stuff. Or you could always go to one of the governments that have researched the use of such compounds as weapons. The Soviet Union certainly did; maybe there's a pile of it still sitting around somewhere. Or you could go to - yes - Iraq. They've looked into it, too.
And maybe there's a pile of it sitting around there as well. And if there is, what are the odds, do you think, that a team of people driving around in conspicuously marked trucks will find it?
Edited later to add a PDF link and the link to Bob Standaert's faculty page at UIC. And - not that he asked me to do so - but I should add the the political comment is mine, not his. I could go on for pages in that vein, but there are plenty of other places to go in the blogosphere for that!
|
Insider Stuff, From Outside
It's been a while since I wrote about Imclone. The ongoing story, as many will recall, has Merck KGaA (no relation to the US Merck) using their drug, Erbitux, in a colorectal cancer trial in Europe. On Friday, the Wall Street Journal ran an interesting piece pointing out that Imclone's stock has run up 43% since the end of January. You'll have noticed that this outpaces the rest of the market a bit.
So what's up? There seem to be rumors flying around that the German data is positive - I haven't heard those myself, but, y'know, nobody ever is kind enough to slip me a phone call with early news on clinical trial data. Were I Martha Stewart, of course, things would be different (and how.) According to the Journal, the workup of the trial data is supposed to be finished this month, and Merck is supposed to make a "brief, general statement" about it all. The details will be in presentations at the ASCO (American Society of Clinical Oncology) meeting in May.
The article quotes a hedge-fund manager who says that an investor relations person at Merck told him that the trial data were positive. (Wonder if they'd tell me the same thing if I called? Time to air out my alarming German! "Ah, entschuldigen Sie bitte, koennen Sie mir vielleicht sagen, wenn Ertibux geklappt hast?") Another hedge fund fellow says that Harlan Waksal (the Waksal of the moment at Imclone) called the trial data positive, too.
Well, that's just fine - if you can get Merck to pick up the phone, or if you can get Harlan Waksal to not hang up on you. But last I heard, there's this thing called "Regulation FD" over here in the States. Many European companies haven't really internalized that one, but American ones should have by now. And I find it really, really hard to understand how Harlan Waksal can tell some fund manager that the trial data are positive without being obliged to tell the world within 24 hours.
Now, a former FDA lawyer is quoted as saying that disclosure might have been taken care of back in December, when Merck said it hoped to file for approval in the second quarter. But I think the stock price action in Imclone speaks for itself - this recent wink-nudge stuff seems to be having a lot more effect on the chart. I mean, everyone says that they hope to file for approval of their drug after the trial finishes. You haven't looked at the data yet, not thoroughly. Why would you say anything different? (The alternative of not saying anything is probably the best one.)
And I'd be awfully careful about characterizing data before it's been completely worked up, too. All sorts of companies (well, OK, small biotechs are a bit over-represented) have characterized their clinical data in terms that weren't quite in focus with reality. The only way I'd feel safe in calling the data "positive" is if it were overwhelming, in which case I'd be keeping my mouth shut anyway.
It'll be a long wait until ASCO, which embargoes presentations. Let's see how tight-lipped Imclone and Merck can manage to be.
|
Copyright 2003-2004 Derek Lowe. All rights reserved. Terms of use
|
|
|
Archives, RSS & Email
|
Click here for access to our archives.
|
|
Dictionary of Scientific Quotations
If the Universe is Teeming with Aliens
(Then where is everybody?)
The Night is Large
A Martin Gardner collection
Eurekas and Euphorias
|
