About Us | Corante Research NEW! | Reprints | Advertise | Contact Us | Services


INDUSTRY NEWS today's headlines

Law & Policy
Personal Technology
Venture Capital


Apple Matters
Broadband Wireless
Customer Intelligence
Get Real

WEBLOG COLUMNS today's highlights

Amateur Hour
the "me" in media

The Bottom Line
the economics of IT

Brain Waves
neurons, bits & genes

nodes & networks

the politics of IP

Got Game?
the future of play

creativity & innovation

The Importance of...
networks and law

In the Pipeline
drug discovery

Living Code
biology & information

The Loom
life, past and future

Loose Democracy
hyperlinked citizenry

social software

Moore's Lore
new technology

Open Mind
open source

Transition Game
sports & tech

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 mugshot

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.

Corante.com Navigation

About Us
Contact Us
Press Coverage

Advertising Info
Corante Research
Buy Reprints
Custom Services

Privacy Policy
Terms of Use

A boy and his chemicals!
Oliver Sacks - "Uncle Tungsten"
(The famous neurologist was nearly a chemist. . .)

Guaranteed 100% Incorrect
The Experts Speak!
(Authoritative and completely wrong)

Paradox and Puzzles
Labyrinths of Reason
(Odd problems, and why they're odd)

The Mind of a Mnemonist
The Mind of a Mnemonist
A true story - he never forgot a thing

Prime Obsession
Prime Obsession
(The mighty Riemann Hypothesis)

powered by WebCrimson

IN THE PIPELINE: drug discovery

By Derek Lowe

Wednesday, July 30, 2003

Bravo, You Dolts

Well, it's finally time for me to comment on last week's vote in the House on drug reimportation (although I don't expect there's much suspense about what my opinion is.) First off, I'm assuming that this bill is going to die in the Senate, since 53 senators have already signed a letter to that effect. But the night is young. There's no telling what my industry's lobbying could manage to accomplish, because they turned the House vote from a setback into a disaster.

I'm not the only person who thinks that way, of course. Let's go to the Wall Street Journal's Alan Murray: "The size of that victory was partly a backlash agains the industry's heavy-handed lobbying tactics. . .the industry needs to stop hiding behind cheap front groups and start telling its story straight." And here's a bit from an excellent roundup in the subscription-only newsletter BioCentury: "The vote was a wake up call that most members of the House are willing to support implicit price controls on biopharmaceuticals. They either do not believe that price controls would stifle innovation, or they do not care. . .the outcome put a bright spotlight on the deep antipathy to the pharmaceutical lobby's attempts to stifle its opponents."

That's the truth, PhRMA guys, and I hope you're learning to live with it. All this "Seniors Coalition" fake-grassroots stuff doesn't cut it any more, and it was idiotic to think that it would. Similarly sparkling was that plan of writing the talking points for the Traditional Values Coalition, in such a way that reporters could see that the software was actually licensed to PhRMA employees. Oh yes, when my mortgage payment rolls around, I can take comfort in knowing that my livelihood depends on people who thought that this was a slick and effective move.

I've said it before: the "unsafe drugs" argument is a loser. It smells, and the smell clings to people who take it seriously (or pretend to.) There are real arguments against drug reimportation, arguments that thinking adults have a reasonable chance of understanding and sympathizing with. But we can't make them while we're pretending that antihistamines from Edmonton are going to poison everyone, now can we?

As I said, I believe that this bill is going to evaporate in the Senate. But what have we here? That letter with the 53 senatorial signatures? The New York Times now has a copy of a letter signed by one of PhRMA's red-hot dealmakers, in which he helpfully points out that they're helping Sen. Santorum with "the logistics of getting this 'Dear Colleague' (letter) out to as many offices as possible. . ." Well, Santorum (who wouldn't have been my pick as point man, exactly) is at least saying that he initiated the letter. But doesn't this give some opportunistic Senator an out, if needed? "Why, I had no idea this letter was being circulated by (gasp) a lobbying group! I'll have to reassess my position. . ."

Oh, there are no limits to what we can accomplish over here in the drug industry. We can stop diseases in their tracks that used to mow people down like ripe wheat. We can bring some people back from the very parking lot of the funeral parlor, and we're staying up late at night trying to figure out ways to do it some more. And we can then take what should be the biggest reservoir of good will around, drain the whole damn thing, leap into the resulting mudhole and sink clear out of sight. Arrr.

posted at 10:16 pm

Tuesday, July 29, 2003

A Shameless Plug

Speaking of the author of BorzoiBlog, I recall that he was editing an American Chemical Society volume a while back, which made an appearance on his blog. Well, I'm not the editor of this one: Oxazoles: Synthesis, Reactions, and Spectroscopy, Part A, but I did write a chapter for it. Mind you, it's the shortest chapter in the book, but if you look at the tables of data, you'll see why I had to draw the line.

At $325/copy, it's the perfect time to stock up for those hard-to-please folks on your Christmas list. Or, for my fellow chemists, it's the perfect time to ask your library people to buy a copy or two. The royalties will enable me to buy not just one, but quite possibly several pizzas. And the long-suffering editor, Dave Palmer of Johnson & Johnson, will be able to do even better than that!

posted at 9:51 pm


I envy Greg Hlatky for his ability to wash his compounds away from most of his impurities. That's a rare event in my hood, although I've had a few systems where I could get away with it. Of course, he's in the organometallic chemistry world, a strange hybrid territory inhabited by people who have really good glove boxes, but who still don't mind seeing their compounds burst into flame now and then.

The last time I did much in the way of organometallic chemistry was back in my senior undergraduate year. I believe that I've mentioned this reaction before, but in case I haven't, here 'tis: You see, my advance inorganic chemistry class had a rather free-form approach to laboratory work. You just had to cover the bases of various lab techniques and types of compounds, and it was up to you to find some good experiments to do it. The professor for the course had to approve your plan, which he did by judging your presumed competence (or your presumed expendability, I guess.)

So that's how I ended up making a compound called "magnecene," more properly bis(cyclopentadienyl)magnesium. The corresponding iron compound is known as ferrocene, and it's a perfectly stable beast, brick-red and crystalline. Bit the magnesium analog is a bit more rambunctious. It reacts with oxygen, it reacts with water, with carbon dioxide, and it's willing to react with lots of other things if given the chance. I think that's why it appealed to me.

And the synthesis had a certain brute-force attractiveness to it, too: Take some magnesium metal. Heat it up to 600 degrees. Blow some cyclopentadiene vapors over it, and there you are. It doesn't get much more elemental than that. I had to rig some things together to get this to work, since we weren't quite equipped for everyday reactions of this kind, but one afternoon I cranked up the heat in a vertical tube furnace and let 'er rip.

I had a distillation setup on the side, because cyclopentadiene doesn't last long as a solution (it combines with itself in a reversible reaction, and you have to crack it back to the monomer.) Blowing nitrogen gas across the top of the still head took some of the vapors across the hot metal, and conveniently kept it from exploding in a shower of blinding white-hot fireworks. Believe me, magnesium metal at six hundred degrees is ready to party.

Things were moving along just fine until I noticed no more vapors coming from the far end of the tube. Deciding that something had gotten plugged up, I turned the nitrogen flow off to keep the pressure from building. I'm still not sure if that was a good move or not. As it was, I looked up at the big elbow piece of Tygon plastic tubing that was holding things together (told you I had to rig this thing,) and noticed that it was now an interesting lumpy black, and making a peculiar sizzling noise.

Yep, my bolus of red-hot magnesium turnings had zoomed back up the tube and had become stuck as they tried to melt their way through the Tygon. And I'm fervently glad that the stuff held, since it would have been pretty festive had some air made it into the system. I eventually made some fine feathery plate-like crystals of magnecene - after replacing the tubing - and went on to a few even more ill-advised syntheses. A mis-spent youth, perhaps. . .I can only begin to imagine what my company's safety committee would make of it if I tried to set up some of these now!

posted at 9:37 pm

Monday, July 28, 2003

Old Friends

I've left myself very little time to blog this evening, for which apologies. I had some extra time taking care of the kids today, with some time off work, and that does vacuum up the leisure. They're getting some science from me (since they're a bit young for Philip Larkin's poetry, y'know.) This evening, for example, I set up the microscope that I was given some thirty years ago, and showed them some of the fauna in the water of their tadpole tank.

They were interested. My nearly five-year-old son saw a large Paramecium-type organism swim by and exclaimed "That one looks like a flounder!" And I was very glad to see that many of the creatures I remembered were carrying on these many years later: Vorticella, Cyclops, Euglena, and others. (I have the same sensation when I look through my telescope. Saturn looks the same as it did when I was eight years old, and the colors of the double star Albireo are just as bright and startling.)

It's odd to think of how long it's been in "protozoan years," but there there everyone is, as if I'd never left. But I had. The house I lived in in Arkansas in 1973 is still there, but the window whose light I used for the microscope is now shaded by trees - which were saplings. Buildings where I went to school have been gutted or demolished, and the whole site of my first job in the drug industry has been leveled (and is now the parking lot of a Home Depot store, of all things.) It's good to know that there are worlds I've visited that haven't changed at all

posted at 9:59 pm

Sunday, July 27, 2003

The Big and the Really Big

A little while back, I had some mail from J. Bowen over at No Watermelons:

When I was in engineering school I noticed that the electrical engineers all wanted to fool with chips rather than big machines and distribution stuff. No doubt that was where the money and snob appeal was, but none of that stuff ran without juice from a big power plant somewhere running down some transmission lines.

I'm guessing that you have a similar situation in your field. The process guys have an art of their own, and certainly they're essential. But no one will ever know their names, no matter how difficult the practical problems they surmount. And without them, well...people don't want to pay what they'd have to to produce drugs by hand in your lab.

Are the process guys a significant technical bottleneck? Should we be directing more talent in their direction?

Well, one problem is that these days, no one wants to pay the price of the drugs that are made efficiently on large scale, either. But he's got a point. Historically, drug discovery has been seen as a bit more glamorous than process work, although the latter is just as vital to a company's success. Good process people get respect inside their own company, but for a variety of reasons, they're often not well known outside it.

The synthetic methods used to makes drug candidates in labs like mine is never good enough to be scaled up to production. There are always refinements that no one at my end would think of or worry about, but matter when you're talking hundreds of kilos. A smaller or less difficult waste stream, tiny improvements in the purity of the product, an extra decimal point of reproducibility: I'm never going to worry about these very much. The largest amounts someone like me deals with are in the low hundreds of grams. And believe me, we hate to work on that scale. The exact analogy is someone trying to feed a 100-person banquet with home kitchen equipment.

It's the folks next to us down the development pathway that have the equivalent of restaurant equipment. Things like larger flasks, more powerful stirring equipment, amd higher-capacity rota-vaps make a big difference in your life on a 500 gram scale, not to mention above. And there are levels beyond this, of course. When you start to get to pilot-plant scale, the gear isn't just hefty versions of normal lab equipment, it's a different beast completely. That's when you start thinking about what floor to run your reactions on, so you can gravity-feed the product down the pipes into the next step.

All this shows up in the "cost of goods" estimate that a company does on a new drug. That takes into account raw materials, labor, time, environmental impact, all the things that add expense to the final product. It's rare for a drug to get stopped for cost-of-goods, although it has happened. The real showstopper on large scale is irreproducibility, or a difficult purification. Process people would much rather have a consistent, reliable product that has to be cleaned up, rather than a route that gives wonderful product once out of every four runs.

So to answer the original question, I wouldn't call process chemistry a bottleneck. Things can slow down a bit at that stage, but not many drugs come to a halt. Compare that to the ones that fail because of unexpected toxicity, or the number that fail because they unexpectedly just don't do anything, and process and scale-up look like a happy island of tranquility. Not that you'd notice much happiness when there's a big run of compound going on a tight deadline, mind you, but relatively speaking.

posted at 9:35 pm

Thursday, July 24, 2003

A Couple of New Ones

I've added two new links at the right. One is Corey Nahman's site, which is a one-stop roundup of pharma news, particularly stories that pertain to marketing. And I've also added Uli Iserloh, a fellow medicinal chemist. His site has a mixture of things, with a section on chemistry and a good amount of science content. But he's apparently brighter than I am, since he's chosen not to talk about the drug industry on his site, though!

posted at 9:46 pm

Applied Voodoo

I have a difficult compound that I'm trying to purify these days. The usual methods someone in my position would use aren't too helpful, unfortunately. That would be some variation on column chromatography, either by hand or through an HPLC, but we've already seen that these compound fall apart under most of those conditions. (You can tell I'm talking about an intermediate here, and not a final drug.) So I'm going back to a classic method, one that goes back to at least the 1800s: recrystallization.

It's common among chemists to moan about crystallization being a lost art, and compared to a century ago, I'm sure it is. Back then, it was the only game in town, so people got pretty skilled. But we still use it, and we can still get it to work when we need it. For those outside the field, it's pretty easy to describe what's going on. You take your compound up in some solvent that will dissolve it when it's warm, and not dissolve it when it's cool. Heat your stuff up, make a clear solution, and sit back and wait for the crystals to start growing and snowing.

Sometimes, anyway. If you're too aggressive with the solvent, or your starting material is too messy, then when things cool down your compound will "oil out," which is just what it sounds like. You get this rain of goo piling up in the bottom of the flask; the only thing to do is start over. And it's easy to mess up in the opposite direction, too, using a solvent that's too strong and never lets your stuff come out again. As my post the other day would indicate, you rarely see anyone use DMSO as a recrystallization solvent, for just that reason.

Actually, we don't use pure solvents as much as we use mixtures. Often the thing to do is use a fairly strong solvent when you heat the solution up, then add small amounts of the poor solvent before it cools back down. Your compound finds itself coming back to earth in a much less favorable mixture than it remembered, and with any luck, will decide the hell with it and come out of solution gracefully. When it works, it's a wonderful thing. All chemists like crystals; never trust one who's indifferent to them.

They come in several different shapes, and with care can be grown with great size and regular features. Thin needles are common, but you can also get chunky prisms or mica-like sheets. The crystal form usually changes depending on what solvent you use, and plenty of other factors are influential, too: stirring (or not,) concentration in the solution, temperature. On one occasion, over twenty years ago now (I was still an undergraduate,) I got a compound that came out as large hexagonal plates with beveled edges. They're still some of my favorites. I can remember the exact structure I was making. . .and it's tempting to make some just for the heck of it so I can see them again! I wonder if I can work this stuff into a drug molecule. . .

posted at 9:40 pm

Wednesday, July 23, 2003

Turbocharging An Enzyme

The latest issue of Science has a report of a potential new kind of diabetes drug. A group at Hoffman-LaRoche has come across an unusual long shot: a compound that actually makes an enzyme work better than it usually does. This is intellectually interesting; but if you read on, you'll hear the case made that intellectual interest is all this compound may have.

Enzyme-targeted drug discovery is mostly concerned with messing things up, because that's what we know how to do. Lobbing a small molecule into an enzyme's active site will tie it up and keep it from catalyzing its usual reaction, but making an enzyme work even better is much harder. Which makes sense, if you assume that the enzyme has evolved to be very good at its job - there's really nowhere for the activity to go but down, in most cases. So down is where we send it. As far as I'm aware, every drug on the market that targets an enzyme is an inhibitor. If you want to make an enzyme work better, you usually have to hope for some other enzyme in the system that naturally regulates the first one's activity. With some luck, you might be able to inhibit that one instead, freeing up your target to do its thing unhindered.

But it's not impossible to make a pure activator, as this example shows. The Roche group was looking for something to activate glucokinase, a key enzyme in the pathways regulating blood glucose. The enzyme phosphorylates glucose, which is general marker of sugar levels. When more of it shows up, the body reacts to a perceived sugar excess: the pancreas secretes more insulin, and the liver stops making glucose from scratch. Not a bad dual effect for a diabetic patient.

The details of how Roche found their compound are interestingly incomplete. The paper says they "screened a library of 120,000 structurally diverse synthetic compounds" to find their initial lead, which may sound impressive, but I have no doubt that Roche has several times that number available for screening. These days, all the larger companies do. This seems to have been some sort of subset, but the selection rules aren't clear. Their initial hit (not shown in the paper) was optimized to give the compound of interest. To my eyes, it's a perfectly reasonable drug-like molecule, nothing odd about it at all. Plenty of uglier-looking compounds are out there on the pharmacy shelves

The compound binds at an allosteric site, a drug-binding pocket somewhere other than the active site of the enzyme. Presumably this is a regulatory site, meant to turn the activity of the enzyme up or down. Binding to such things isn't a high-percentage play in drug development, but it can be done. The Roche team shows impressive effects on the enzyme in vitro, and on rodent models in vivo. In fact, they show effects in a strikingly long list of rodent models. You can tell that they were really trying everything they could to convince themselves that this was a real and robust effect. If I'm not mistaken, they refer to at least ten different strains of mice and rats, in all different sorts of experiments.

So it's safe to assume that they've been working on this compound for quite a while, and that we're only now hearing about it. Support for that guess comes from the list of authors, two of whom have already left Roche for other companies. There's always a background level of turnover in the drug industry, but this does make me think that the paper's been brewing for a long time.

Is it a drug? That's a very good question. There's a lot of potential here, clearly, although it's possible that this sort of therapy could reduce glucose levels too far. You'd want to check the size of the therapeutic window carefully. As usual, it's impossible to say just what Roche is up to, just as it's impossible to say what any drug company is really up to. They could be well on their way into the clinic with this compound, or an unnamed successor. I hope so, since a new diabetes treatment option would be worth having.

Or there could be another explanation. Often these off-the-beaten-path therapies don't work out, and the only thing they're good for, frankly, is to make a big splash in Science. There's a very similar example from the diabetes field, actually. A few years ago, Merck announced that they'd found a molecule which reacts directly with one key binding site on the insulin receptor. (Not the one where insulin binds - it's hard to imagine a small molecule doing much good there - but another kinase activity.) This was a first, and it got the appropriate publicity. But the molecule itself was rather hairy. In fact, Merck scientists usually found themselves at meetings saying things like "It isn't really as toxic as it looks." Although they've spent a lot of money and a lot of time on it, nothing appears to have come out of the effort, unfortunately. And if something does, it's surely not going to be with the molecule that got all the attention. If that one were working, I think we'd have heard more about it.

So is this the middle of Roche's adventure with glucokinase, or just the beginning of it, or was the end already reached a year ago? There's no telling. Sometimes the industry reminds me of a gigantic game of stud poker. . .something like five-hundred-and-seventy-three-card stud. . .

posted at 9:13 pm

Tuesday, July 22, 2003

Some More Things I Need

I have a chemical intermediate sitting in my hood that I can't wait to use. But I'm going to have to, because I used the Wrong Solvent in the reaction that made it. Well, not exactly the wrong one; I'll explain, for the benefit of the non-chemists in the audience.

Solvents have all sorts of properties, well-known to those of us, uh, skilled in the art. Bench chemists can pick from things that dissolve almost everything (like DMSO) to things that dissolve, well, not much (like hexane.) In between you have ether, ethyl acetate, methylene chloride, workhorses all of them, and some other like toluene and acetone that see some use, too. Each of them has its own personality, and thus its own quirks. For example, toluene will take water out of your reaction when you evaporate it, which can be a nice bonus. You can't expose methylene chloride to sodium metal, and you can't use iodine in acetone. Both combinations will react, and can liven up a dull afternoon, for sure. (To be fair to methylene chloride, there are a lot of things that don't play well with metallic sodium.)

But it's a fact of life that the ones that are the most powerful solvents are the hardest to deal with. Take DMSO. If warm DMSO won't dissolve your compound, you might want to think about working on something else (or you might want to make sure that you didn't weigh out some sand by mistake.) But the same properties that make it grab onto your molecules make it hold onto itself, and that gives it a high boiling point. We like the opposite: solvents that evaporate easily, because you can just boil them right off in a rotary evaporator under a bit of vacuum. If you're going to try that with DMSO, you'd better have something to read, because it's not going anywhere fast.

One way to lose it is to dilute everything up in another solvent and wash that with water. DMSO tends to stay in the water layer, so a few rinses will generally get most of it out. Of course, if your compound likes to go into the water layer, too, like mine, then that's not going to help much. Or you can put the whole flask on a high vacuum line, which is stronger treatment than most rota-vaps. That's OK for small quantities, but starts getting impractical. Like if you forget to maintain your trap in the vacuum line and fill your pump up with DMSO, which doesn't improve it much.

There are other solvents in the painful-removal category, like dimethylformamide (DMF.) Many's the time I've cleaned up a reaction, put the final solvent layers on the rota-vap, and come back to find a lot more stuff left in the flask than I was expecting. Then I remember - oh yeah, I had to add a few milliliters of DMF to this thing. And there it is again, wagging its tail and ready for more, faithfully clinging to my compound.

So, on my wishful-thinking wish list is a solvent that's as powerful as DMSO, but evaporates like ether. OK, maybe not like ether; that one's a little too eager to take off and wander around. Ethyl acetate, maybe - it's better-behaved, and smells better, too. I'm not sure how I'm ever going to get that combination, though. t's like asking for something as light as popcorn but as strong as titanium. But a man can dream. Even if it's just about solvents.

posted at 10:01 pm

Monday, July 21, 2003


Today's Wall Street Journal has a ringing editorial on drug reimportation. It won't change anyone's mind. This is one of those issues you have a strong opinion about if you've bothered to have an opinion at all. Not many editorials change many minds about anything, to be realistic about it, but the chances here are particularly slim. I mean, just check out some of the hair-tearing rhetoric from Congress; it makes me feel as if my job title should be "Rapacious Scum." That'd print up as a memorable business card. . .

I did enjoy the editorial's second paragraph. A note to my industry's lobbying organization, PhRMA: guys, when you've lost the Wall Street Journal's editorial page, you've lost everyone:

We don't have much time for the pharmaceutical industry's contention that such imports would pose a significant health risk to American consumers. But sponsor Gil Gutknecht (R-MN) is selling nonsense when he argues reimportation is about giving U.S. consumers access to "world market prices." Rather it's an attempt to import foreign price controls. At best it won't work. At worst, it will threaten pharmaceutical innovation, which is increasingly done by U.S.-based drug companies.

(Of course, as the Manhattan Institute's Robert Goldberg pointed out last week, this importation of cheap goods can go too far. If you don't believe him, why, just ask Gil Gutknecht! When it comes to milk products, he's all for keeping U.S. prices nice and high - Minnesota has a dairy industry, you know. Steel, sugar, what-have-you: all sorts of goods get the high-price treatment from Congressional admirers of cheap pharmaceuticals. It seems we can't possibly get overcharged for those other things. Just drugs. Who knew?)

I particularly like the emphasis that reimportation advocates put on what they're pleased to call "market prices." You'd think that in the rest of the world, drug expenses have just naturally drifted on down to comfortable levels which make everyone happy. How to explain high prices here, then, except by invoking ravenous greed?

Well, I won't spend too much more time on the subject, since I'm sure that my readers will have seen all my points polished to a gleaming shine already. It's good to know, though, that the Journal finds PhRMA's safety arguments ineffective, since I've ranted about them several times myself. ("Ineffective" is a polite way of saying "stupid and lame." I'm trying to teach my kids to be polite, and I need to lead by example.) But this is the battleground that the industry has chosen, so we may just be stuck with it.

Now the proponents of reimportation just have to show that Canadian supplies are as free of contamination as those in the U.S., and they've won the point. By then, pulling out the real arguments (see the above) will just make the pharma industry look desperate and shifty. And I'm having trouble picturing the upside to looking desperate and shifty, but I'm sure that those go-getters at PhRMA have a wonderful plan for that contingency. Ahem.

posted at 10:07 pm

Sunday, July 20, 2003

Age, And the Only End of Age

No sooner do I link to an article about people gulping down things like Ritalin then the news comes of its use in treating the elderly. The Wall Street Journal gave this story some play on Friday, and it's an interesting one.

The stimulant could be useful in treating general depression and lethargy in older patients. That's a well-known problem without any well-known solutions. It complicates treatment of other diseases, because the affected patients are too apathetic to do much about drug or physical therapy. It could be that Ritalin (methylphenidate) might not be the best thing for this. The New York magazine article seems to indicate that there are other favorites (such as newer drugs like Adderall) among the cognoscenti. And treatments in elderly populations are notoriously hard to generalize. But there could be something here, and it raises the better-than-well question again. After all, what is the normal condition of old age? Cheerfulness and initiative, or not?

I wrote about this in the early days of my Lagniappe web site, this one's predecessor:

As I mentioned previously, I've been reading the letters of both Kingsley Amis and Philip Larkin. One thing you notice in any Collected Letters book (try Evelyn Waugh's) is old age creeping up on the writers. It's less noticable in Larkin's case; his personality famously made him sound about 70 years old for decades. But with Amis, it's clear that you're reading the thoughts of a young, a middle-aged, and then an old man. This process is chronicled from just outside in his son Martin's book Experience.

Don't get me wrong - Amis's letters are wonderful, even the late ones. But a sort of hardening of the personality takes place, kin to atherosclerosis, and it's a common thing to see. What I wonder is how much is due to just plain experience and weariness with the world (seeing the same mistakes being made the same ways, again and again,) and how much has a neurologic base.

What would Philip Larkin have been like on Ritalin? Perhaps he wouldn't have stopped writing when he did - his last years were famously unproductive. But then again, the man who wrote "Deprivation is to me what daffodils were to Wordsworth" would probably have produced a very different body of work. Assuming, of course, that he produced anything. Perhaps he would have just cranked along amiably as the head librarian at Hull, well-liked by all. Or perhaps, if he'd started on the drug early enough, he would have been a novelist, which was his early inclination. Or bagged the literature completely and become an art dealer, or owned a restaurant. "Some mute inglorious Milton here may rest," if I can continue quoting English poetry. Good thing, or bad?

posted at 9:26 pm

Thursday, July 17, 2003

Obvious to One Skilled in the Art

This will be a short post today. I spent a good part of the day proofreading a patent application, which always erodes my consciousness a bit. I spent a good amount of time on the experimental section, looking for problems (it was mostly clean.) Then I waded into the claim language for a while, until I felt it drying out my cerebrospinal fluid (well, I think that's what it was,) which made it unsafe to continue.

I'm an inventor on this one, thus the proofreading. Well, I should revise that: I believe myself to be an inventor, but the patent lawyers will have the final say on that when we file. You have to have made a real contribution, something in the patent that you can point at and explain. Working on someone else's idea doesn't count, either, no matter how much wrestling it took to realize the thing - that's a lesson that industrial chemists learn very quickly. For us, the list of patents on someone's resumé is the most impressive part, the one that best summarizes their accomplishments in the field.

If I have to, though, I can analyze a patent claim pretty well. I need a clean sheet of paper or a blackboard, and I have to close my office door and take the phone off the hook. Interruptions really throw you off when you're holding these subset-of-a-subset-of-a-conditional-group thought chains in your head. Patents have to disclose things, and they have to have explicit intellectual property claims, but they don't have to do either one gracefully. Many are written in a deliberately snaky style, with key sections deliberately parked in obscure places. What it must be like to be a patent examiner in this field, I have no idea. And here's hoping I never find out!

posted at 9:55 pm

Wednesday, July 16, 2003

Better Than What Passes For Normal

A colleague pointed me to this fascinating three-part article from New York magazine on "cosmetic pharmacology." It's been clear for some time now that there are plenty of people taking plenty of mood-altering prescriptions, but this piece really brings that trend into focus.

Some disclosure is appropriate here; bear with me after I've made it. (It's not about how I work for a drug company; that one's pretty well established.) No, here's what I need to reveal: I have never sought out a mood-altering drug. Probably the closest I get are foods that I like, things that make me feel good when I've eaten them. But I don't drink or smoke, and never have. I've never taken a CNS-active prescription, and I've never used any illegal drug of any kind. I realize that I sound like I'm being vetted for some sort of political nomination, but for this subject I think it's useful to know my default settings.

Well, since it's a New York article, it automatically has a generous dose of only-in-New-York-ism. You have to strip out all the everyone-here-is-uniquely-stressed-and-in-a-hurry stuff, which is a line I've never bought, even when I lived within sight of the NYC skyline. That done, what remains is a remarkable view of a large class of people who are taking pretty much everything that's on offer. antidepressants, anxiolytics, pain medication, what have you.

There have always been people like this. But I think that this is a different (and larger) group. Twenty years ago, most people with such self-medicating personalities were at greater risk of ending up in pretty bad shape. The drugs involved were less selective, for one thing, and there was more overlap with the use of (generally more dangerous) illegal drugs. But according to the author, Ariel Levy, New York drug dealers routinely see more demand for the prescription medications. I assume that's partly due to shorter supply, but there's a real demand component, too. With the stigma disappearing and the options increasing, a wider population than the needy, desperate, or thrill-seeking is trying things out.

But, to my surprise, I can't decide just how bad a thing that is. Don't get me wrong: by my standards, many of the people in this article are completely off the chart. But is that just because I have a (relatively) even-keeled personality, and have never felt the need? What fraction of the population would benefit from taking, say, SSRIs? And of that group, what fraction actually takes them? I won't pretend to know the answer; we've been fighting this out since Prozac first came on the market. There's a real chance, which can't be ignored, that the beneficial market for such things is very large indeed.

Of course, we could still be setting ourselves up for major problems. No one knows the effects of taking an SSRI drug for twenty-five years. They just haven't been around that long. Similarly, no one how taking one at a relatively early age might affect you when you're sixty years old. No one's made it that far yet. And the long-term (even the short-term) effects of all the combinations that people are using? No clear idea yet, either. As someone who spent eight years doing CNS drug research, I can tell you that we're not going to be able to even make reasonable guesses. The brain's too complex (it's damn near too complex for CNS drug research to exist at all.)

And there's another problem that the article addresses near the end. Says one interviewee, who doesn't seem all that calmed-down to me:

"If you're taking something to make your life suck less, then why don't you just make your life suck less? This is 2003 in an advanced society of which you are one of the most fortunate members! Think if you were a little child in Cambodia who never got education, and compare that to your incredible fucking life! And you're depressed? How can you be sure it's not just that you're a spoiled brat? They say chemicals are not something that you would respond to if you were not depressed, but part of me thinks that's bullshit."

There's a good point or two in there. I think that people who are taking mood-altering chemicals because they don't like their marriage, or their job, or their apartment are making a mistake. The risk-benefit ratio doesn't come out right, as far as I'm concerned, not to mention that the prescription treats the symptoms and not the problems. But if you have one of those incredible, uh, procreating lives the person above speaks of, and you're still anxious and depressed all the time, what does that mean? I've no time for most explanations from psychotherapy, and I don't have a lot of faith in the ability of most adults to make significant changes in their personalities, either. So what does that leave?

A prescription, presumably. Which is an odd place for me to end up, given my views. But we're all going to end up in even stranger spots in the coming decades, as we start to face up to more and more "better than normal" possibilities in medicine. I've written several times here about life extension research - and if that doesn't fall into the category, what does? The idea of a substance that gives me another twenty healthy years excites me, but I'm still nervous about the idea of one that makes me happier. Cartesian dualism dies hard. What will replace it?

posted at 8:49 pm

Tuesday, July 15, 2003

Will That Be Feast or Famine, Sir?

Yesterday's post-ending quote gets to a constant problem in the drug industry (or, for that matter, in any industry that depends on scientific research.) The business side and the research side of the company are trying to do two different things.

Over in the business world, management would like solid, smooth, steady performance. (A little better each year, please, while you're at it. The stockholders like it that way.) No nasty downside surprises - and if the price you pay for that is getting rid of any upside surprises, well. . .sold. You've got a deal.

Everyone sets performance goals, using things that can be measured ("deliverables," they're called by some) and everyone gets rated according to whether they meet those goals. And it's assumed that your progress toward these goals largely depends on how hard you work.

I've long thought that you have to work in research before you truly understand how incompatible all that is with the usual course of science. We do not deliver steady performance over here. A research program is a recipe for a jerky, jolting ride. It features long stretches where you'd swear everyone had died or moved away, and no one was ever going to discover so much as a recipe for cheese dip ever again. Those lulls are seasoned with short periods where dizzying, inexplicable results pile up faster than anyone can interpret them. And the reason we bounce around like tennis balls in a clothes dryer is because, by business standards, we have no idea what we're doing.

No, really, we don't. We don't know how to generate more ideas for projects, and we have had one heck of a time making those projects fail less often. If we knew how to make drugs that worked, we'd be making more of the things. Especially right about now. (Reminds me of the old Orson Welles wine commercials, with the tag line "We will sell no wine before its time." Those always made me picture some green-eyeshaded guy in the background with a calculator, looking up and saying "Whoa. It's time! Oh yeah, Orson, it's definitely time. . .")

So the two halves of the business don't exactly feel comfortable with each other. The business end can't understand why research won't set and meet its performance targets, like everyone else has to. And the research end despairs of the business end for even asking.

And these performance swings can hold all the way down to individuals. Come performance review time, one of the things you can reliably get downgraded for is inconsistency. Up and down periods, good and bad years don't make you look very effective. But some of that is in the very nature of the work. There are going to be projects that don't allow for much productivity, and others where you can just crank away. (And some of it is just in human nature, or at least mine. I have periods when I'm much more productive than others.) A key to surviving, not to say prospering, in an industrial research organization is to figure out how to even out some of those project and personal sine waves.

But the big ones refuse to be evened out. Every drug company, as far as I can tell, wants to have a steady flow of ideas, of new projects, of clinical candidates. But no one has as steady a flow as they plan for, so back go the planners to those drawing boards you've heard about. . .

posted at 7:43 pm

Monday, July 14, 2003

Some People Can Do What They Like

Physicist Chad Orzel over at Uncertain Principles has some thoughts about problem-solving in the lab:

Back when I was a post-doc, I used to have the occasional arguments with my boss regarding the proper approach to intractable problems. He was firmly of the opinion that there was no problem that couldn't be solved by just staying in the lab continuously until it was solved, while I tended to disagree. After a few hours of futility, my brain just shuts down, and I stop being able to think of new approaches to the problem. Once that happens, unless the problem in question can be solved by rote, mechanical manipulations, I'm useless.

I've always found that when I'm stuck on some problem that just won't give, the best course of action is to set it aside for a few hours, and start fresh after a break. At least half of the major lab problems that got solved in my post-doc days were solved when the solution occurred to me on the walk home after shutting down for the night, or on the walk back in the next morning. Most of the time, I wanted to kick myself for having missed something that seemed so incredibly obvious after a few hours of sleep.

That's been exactly my experience. Solutions hit me at the oddest times, but never when I'm staring at the bench trying to make one appear. I'd say that his former boss's method is probably the least likely to solve really interesting problems. They're just too big and convoluted to attack by brute force. (I'd be willing to bet that he once solved a problem that way, though, and has mistakenly sworn by the technique ever since.)

And that gets back to a favorite topic of mine: what sort of problems one should be working on in a research lab, and the best ways to sneak up on them. I'll quote again from Robert Root-Bernstein's Discovering:

Arthur Hadley, a former president of Yale, observed that in his opinion the best work was "done by men who were a little lazy and had good consciences. Such men would not work for the love of being employed, often in time-consuming endeavors upon subjects which were not worthwhile, but if after due meditation they saw a thing was worth doing, their conscience would drive them to make the effort and accomplish the task. These did the greatest work."

That makes me feel better, I have to say. I'm not always a ball of fire myself, and that description fits me pretty well (up the the "greatest work" part; the jury's still out on that one, and not likely to show up for a while.) I'll close out with a quote from Rayleigh's biography of physicist J. J. Thomson. He's speaking on a point that I'll be coming back to in a separate post, because it has a very strong application to pharma research:

"If you pay a man a salary for doing research, he and you will want to have something to point to at the end of the year to show that the money has not been wasted. In promising work of the highest class, however, results do not come in this regular fashion., in fact years may pass without any tangible results being obtained, and the position of the paid worker would be very embarrassing and he would naturally take to work on a lower, or at any rate, different plane where he could be sure of getting year by year tangible results which would justify his salary. The position is this: you want this kind of research, but if you pay a man to do it, it will drive him to research of a different kind. The only thing to do is to pay him for doing something else and give him enough leisure to do research for the love of it."

posted at 9:29 pm

Sunday, July 13, 2003

Vox Populi

Longtime reader Joe Cerro sends along an interesting item. The Wall St. Journal Europe recently had an article (subscribers only) on a European survey commissioned by the drug industry. The results were. . .interesting, in a sort of hand-wringing way.

The part I most enjoyed was a question about who gets credit for new medicines being discovered and sold. The drug industry gets a good vote there, which is a relief. But the governments involved made a strong showing, too, as did (hold on for this one) pharmacists, of all people. That one puzzles me quite a bit - perhaps people have more of a memory of compounding pharmacies there?

59% of Europeans aren't buying the argument that price controls are bad for the drug business. At one point, respondents had to rate 35 assertions for their believability, and the price-controls-affect-research one was at the bottom. (I'd like to know what the other more persuasive ones were. . ."My alarm clock makes the sun rise," maybe?)

This is a real problem for the pharma companies. I doubt if this survey told people too much that they didn't know (or strongly suspect,) but it's worrying, just the same. The worry I have is that it's too easy to believe some of the things the European public believes. (It feels good to think that some of these things are true! Why not just run with 'em?)

What I'd like to know is what the answers would be for the same survey, given here in the US. It wouldn't surprise me if the drug industry hasn't run that one already. (A control group - we love control groups.) What percentage of Americans think that pharmacists deserve credit for the discovery of new drugs? More unnervingly, what percentage think that legally mandated low, low prices wouldn't have an effect on research? (Now that one I know is being polled. . .!)

posted at 9:16 pm

Thursday, July 10, 2003

Short Topics

A few short items to close out a week of longer-than-normal posts:

Schering-Plough (of whom I am a shareholder) came out with lower profit forecasts, yet again. I've lost count of how many times the company has said that, darn it all, they've just been taken by surprise at how quickly sales of Claritin tanked when the stuff went off-patent. The mystery is why the company thought that this would go differently from any of the other recent patent expirations. I'm willing to bet, though, that some sort of groupthink illusion was at work. It was probably just not done to put up an internal estimate as bad as things have turned out to be: It simply Wasn't Going to Happen; Everyone Knows That and we've been over it before, so let's move on to the next item. Anyone who's worked at a large organization will have seen this sort of thinking in action. Come to think of it, it doesn't take a very large organization at all. . .

Novartis's new Cambridge research center, about which Richard Gayle and I wrote a few items, is staffing away. They're probably getting applications from former employees of Millennium and Vertex, two of the area's more prominent small-to-medium sized operations. Both companies have announced cutbacks in their discovery research, because both are experiencing growing pains. Millennium (whose cuts are more severe outside of Cambridge) now has something on the market (Velcade.) And now that they're no longer a no-product company, they seem to be feeling the pressure to not be a one-product company. And Vertex seems have felt that they're spread too thinly, and are throwing some research areas over the side.

Finally, for those who've wondered about my ugly chlorosulfonic acid reaction that I mentioned the other day. . .well, it's not so ugly any more. The crystalline material that came out of the reaction actually looks a lot better than the starting material, weirdly enough, considering that it went through that used-motor-oil stage along the way. But, unfortunately, it turns out to be the starting material itself, having survived the reaction completely unscathed. How anything can come unchanged out of hot undiluted chlorosulfonic acid, I ask you? Anyway, it's not a purification procedure that I can really recommend. Maybe the starting material is clean enough now to react in what will surely be an even less aesthetically appealing repeat of the reaction.

posted at 9:04 pm

Wednesday, July 9, 2003

Alzheimer's and Insulin?

There's a neat paper in the latest issue of Neurology, summarized in the latest issue of Science (for the many, many people who don't keep up with their subscription to Neurology.) It's a good example of the tricky connections to be found in living systems, and how You Never Know what research in one area is going to do to another.

I've spoken before about the beta-amyloid theory of Alzheimer's which ruled the earth when I worked in the field ten years ago, and continues even more strongly today. One of the big questions, still unanswered, is: where does all that beta-amyloid come from, anyway? On level, we know the answer - it comes from amyloid precursor protein, APP. The hunt for the enzymes responsible for carving beta-amyloid out of APP has finally ended within the last few years. There's still a lot of arguing to do, though, about what those enzyme are supposed to be doing normally, and whether or not they somehow get more active in people who are developing Alzheimer's.

Cut to the jungles of endocrinology. A connection between Alzheimer's and diabetes has been building for some time, but it's been a frustrating business. For one thing, there are a lot of insulin receptors in the brain, and no one's quite sure what they're doing up there. Insulin seems to have effects on cognition and memory, outside of its sugar-mobilizing role, but it's very, very hard to nail that stuff down. Type II diabetics develop Alzheimer's at about twice the rate of age-matched controls, which is suggestive but hardly definitive. (Patients with longstanding Type II diabetes generally have all sorts of other problems, which makes a connect-the-dots approach very risky. For example, their lipid and lipoprotein profiles are all out of whack, and there's another well-known - but not well worked out - connection between Alzheimer's and lipoproteins. And diabetic patients tend to have circulation problems, which can't be helping the brain out much, either.)

Meanwhile, out on the fringes of the diabetes world, there's an enzyme known as IDE, for Insulin-Degrading Enzyme, which is exactly what it does. (Biochemistry resembles the German language in its dogged literal-mindedness of phrase.) On and off, people thought about inhibiting it as a way to improve diabetes. (That's a standard pharmacologic ploy: to get more activity out of a biomolecule, find the thing that clears it out and try to gum that up.) The problem is, IDE degrades quite a few other proteins, many with substantial biological activity, and it's never been clear just what would happen if you stepped in and inhibited it. Plus, there are other ways that insulin is broken down. This combination has meant that IDE has never been a front-line diabetes target.

Back to Alzheimer's. Dennis Selkoe's lab at Harvard has been an Alzheimer's powerhouse for many years now. They spent several of those years looking to see what enzymes might degrade beta-amyloid. That's an interesting approach - the reasoning was that the accumulation of beta-amyloid might be due to a problem with an enzyme that should be clearing it out (a perverse parallel to the drug target strategy I mentioned above.) And, as fate would have it, the enzyme with the greatest activity was IDE. There are surely others, but it seemed to be a major player.

Time for the in vivo data. Just in the past few months, reports have been published from Harvard and from Southwestern/Dallas on mice that have had their IDE enzyme knocked out or severely impaired. And the hypothesis made it through: such animals tend to accumulate much greater amounts of beta-amyloid, depending on how much IDE activity remains.

The Neurology paper takes the big leap to humans - no, not by knocking out their IDE enzyme. For better or worse, we don't know how to knock out human genes yet, although I'm sure there's an instant Nobel waiting for anyone who figures it out (along with a lifetime of head-banging controversy.) What this joint Washington/Wisconsin group did was to infuse insulin into volunteers, along with dextrose so their blood sugar didn't tank, and then take some cerebrospinal fluid. (There's a lumbar puncture hiding in that last phrase; my hat is off to the people who volunteered for this one.) They checked levels of beta-amyloid in the CSF, and found that it increased with increasing insulin, but the effect was only significant in the patients who were more than 70 years old. One hypothesis is that excess insulin is soaking up the capacity of IDE, allowing beta-amyloid to start to accumulate.

I find it remarkable that such an effect can be demonstrated over such a short time. Makes you wonder what the statistics would be like over a longer time period, not that any 80-year-old would (or should) put up with a euglycemic insulin infusion for that long. Another remarkable thing was that they checked the patients with two memory tests (recall of details from a short passage, and a color/word association test.) Insulin enhanced memory performance, which is an effect that's been seen in many other studies (it's one of the poorly explained brain effects I mentioned at the start of this post.) But this enhancement was much less noticeable in the patients who showed elevated beta-amyloid levels in response to the insulin.

Now that's odd. To their credit, the authors make no pretense that it's anything else: "The significance and cause of this unexpected finding remain unknown, and any potential explanations must be considered speculative at the current time." That's about as complete a hands-thrown-up gesture as the traditions of the scientific literature will allow. It's hard to imagine that the neuronal damage associated with beta-amyloid is acting on that sort of time scale. There's clearly a lot of interesting stuff to be worked out here.

This will keep people busy for a while. What's that instant memory effect, and where's it coming from? (And while you're at it, figure out what insulin is doing up there, too.) Is there a potential Alzheimer's therapy here, involving strict control of insulin levels? And before we get too worked up, do people with genetically defective IDE have a greater incidence of Alzheimer's, anyway? What about people who get Alzheimer's with no apparent hyperinsulinemia? Related to that question, how many other enzymes are involved, and what are they? Any better therapeutic prospects?

While I'm on those last two points, here's another question, which I just thought of myself (although I'm sure it's occurred to those in the field): if hyperinsulinemia puts you at risk for Alzheimer's through saturation of IDE activity. . .would some hypothetical non-insulin-dependent mechanism for lower IDE activity mean that insulin levels in the brain are raised? If so, does that compensate for the memory effects of the disease process, or not? That'll all depend on how much non-insulin-associated Alzheimer's is associated with some other amyloid-clearing enzyme. . .

posted at 9:56 pm

Tuesday, July 8, 2003

The Right to Know. . .What, Exactly?

In the early 1990s, I was working at a large drug company in New Jersey, not that that narrows it down very much. A law had been passed in the state legislature called the "New Jersey Right To Know Act," which very shortly began to affect our lives.

This masterstroke required (among many other things) that every stored chemical be labeled with its contents, including the Chemical Abstracts (CAS) number for easy reference. Sounds fine - in fact, it sounds like most things would already comply. But a closer reading of the law showed that there wasn't a clear minimum for what containers needed to be labeled. And there wasn't a clear minimum for the trace constituents in their contents, either. What there was, was a clear schedule of the fines that you could pay, per violation. No ambiguity about those

Our company's regulatory department came up with some fibrillation-inducing worst cases, based on some walk-through inspections of our labs. Their recommendation was that every container of everything be labeled. Just to be sure. So we embarked on a grumbling, grudging labeling frenzy. It turned out that all our older reagents had labels without CAS numbers on them, so that had to be fixed. And any special mixtures made there in the lab had to be labeled. So on went the CAS numbers for sodium chloride and water onto the saturated brine that we use to wash organic solvents in separatory funnels. On went the CAS numbers for water and sodium bicarbonate, for sodium hydroxide or HCl, on all the diluted mixtures that chemists make for day-to-day use.

But it didn't stop there. Every time you asked the regulatory folks if something should be labeled, the answer came back "Yes." Just to be sure. Eye wash stations, safety showers? CAS number for water. Water baths? The same. How about the water fountains out in the hall? Well. . .there's nothing here that says that they shouldn't be labeled. . .and so the madness went.

And the no-minimum clause on the constituents was a real problem. The only escape hatch was that you could label things with the main components and then say "contents partially unknown." It wasn't long before the major chemical suppliers, while rolling their eyes and bumping up their prices, ended up slapping little sticky labels onto every bottle of every reagent. They said "For the purposes of complying with the New Jersey Right to Know Act, Contents of This Bottle Are Partially Unknown." This lasted until the law was clarified, which took a while. I can still find reagent bottles in my lab's cabinets that have the stickers on them.

The whole time I was helping to label water baths and jars of table salt, I kept hearing an imaginary sound bite from some state political campaign: "And I fought for tough new laws, that make the big polluters tell us what chemicals they have, how much they have, and where they keep them!" (applause)

I've been in other labs where other requirements intruded. One workplace had a form attached to every waste solvent container. We were, in theory, supposed to write down the volumes of every solvent that we tossed in there. But the forms weren't up to the task of documenting 20 or 30 liters of mixed solvent, much of which was added in 50 or 100 mL lots. So we ended up consolidating the amounts, and that led to - well, to just guesstimating what a typical load would be. When it came time for them to haul off the waste cans, we'd sit down and say "Hmmm. . .there's probably six or eight liters of THF in there. . .bound to be a lot of ethyl acetate, and hexane, too. . .I think I put some DMF in there, let's call it 300 mL, and a nice round 700 mL of toluene on top of that makes it come out nice and even. . ."

So you can imagine how much those forms were worth. I wonder who was keeping track of the totals, and how much faith they had in the numbers? Sasha Volokh, of Harvard and the Volokh Conspiracy, has written a comprehensive and interesting paper on Right-to-Know regulations, and he addresses this problem head-on:

Every article about environmental right-to-know laws. . .pays lip service to the problems in the data, and then blithely uses the data anyway as an acceptable proxy for risk, harm, pollution, and the like. It is time we took the problems seriously and stopped reading meaning into mandated statistics that at bottom mean very little.

He has some points that directly bear on the proposed EU regulations that I wrote about yesterday, too:

. . .These problems are compounded by the move toward mandating information on chemical use, not just on releases, pollution, or harm. Regulation has traditionally been justified by the need to prevent individuals or corporation from imposing unacceptable risks or harms-"externalities"-on customers, workers, or surrounding communities. Environmental regulations are not always well-advised or optimally efficient, but they have typically been aimed at preventing identifiable harms; activity that does not cause harm has been allowed to remain within a company's sphere of private autonomy. Laws that focus on a company's internal chemical use, without a showing of risk or harm, are inspired by a general distrust of industry and/or try to indirectly reduce the use of materials in general.

Preach it! I think this is exactly where the proposed EU rules sprout from. They make the New Jersey laws, even in their most unworkable phase, look like harmless bunny rabbits. Compared to them, the proposed EU regulations have fangs and breath fire. It'll be interesting to watch various European governments trying to stop or amend these proposals, at the same time as some people here urge that we follow Europe's lead and adopt them as soon as possible.

posted at 9:28 pm

Monday, July 7, 2003

Outbreak of Good Sense May Be Contained, EU Says

And while I'm on the subject of EU regulations, Stephen den Beste's weblog recently mentioned one that's been a hot item for a couple of years now.

What the European Commission is proposing is a huge new regulatory world. It's full of phrases about "stimulating innovation and competitiveness in the chemical industry," but how it can possibly do that is a real stumper. It'll certainly stimulate innovation in bureaucratic paper generation, and competitiveness in the field of which companies can leave Europe first, but I don't think that's what they're talking about.

The proposal shifts the burden of chemical safety right back to industry, which I'm sure is politically popular. But it would seem to make companies responsible for proving that things they use is safe, before they're allowed to use it. That's where things get ugly. As den Beste points out, "safe" is slippery term. Industries use things all the time that are not "safe" in any meaningful way.

(For example, just today I used a reagent called chlorosulfonic acid. It's vicious stuff. You can't even take it up in the usual plastic syringes, which are usually fine for all sort of nasty reagents. It eats them. It softens the barrel in the time it takes you to get the syringe over to the flask, and when you push down the plunger, like as not it just goes "splorp" out the sides like some sort of demonic party favor. Chlorosulfonic acid fumes, hisses, and spares not when it contacts most any organic compound. When I left work, my reaction was thick and reddish-black, not that I expected anything more appealing. What it would do to my hand is anyone's guess, but it wouldn't be an improvement. Is this a safe reagent? Only if it's safely handled, and treated with respect and attention.)

Safety aside, every chemical that's produced in over 1-ton lots in Europe would have to be registered. That's about 30,000 substances, and readers are invited to imagine the time and paper that this process will consume. Furthermore, the estimate is that at least 5000 of those compounds would have to be "evaluated" by the authorities. That evaluation would use data provided by industry, and I'm sure that here is where things will come to a squealing halt. How many rounds of argument can there be over what data are generated, how they're generated, and how they're interpreted? Many, many. Even as it is, every time the chemical industry generates its own toxicology data, half the intended audience doesn't believe the numbers. This situation is going to improve?

There's more. A smaller list of chemicals would require "authorization" before they could be produced or sold. These are presumably the nastiest substances, but who gets to decide what goes on this list? That's sure to be a fine football to kick around. The mechanics of this authorization aren't clear to me, but the thought has all the chemical and pharmaceutical industry of Europe in a panic.

I forgot one of the best parts. All these lists are just for "existing substances." The cutoff is 1981. Anything that's come on the market since Blondie's Autoamerican is defined as a new compound, and is subject to thorough testing and regulatory review before marketing in any amount over 10 kg. That should fuel up the mighty engines of innovation, all right. Just don't come up with anything new, and you'll be fine. (At least my old pal chlorosulfonic acid will escape - it's been reliably nasty for decades now.)

It's become clear that this process is going to be open to all sorts of abuse, and could quickly turn into a license to cripple the European chemical industry. This proposal has actually been too much even for various European heads of state. Gerhard Schroeder vented on the subject recently, sounding like as much like someone from the Cato Institute as I think he ever can. Perhaps this sort of thing can, ahem, "heighten the contradictions," as a once-popular European political philosopher once put it.

(Tomorrow: my enjoyable personal history with chemical regulation. More fun than it sounds.)

posted at 9:58 pm

Sunday, July 6, 2003

A Sporadic Outbreak of Sense

Now this is an interesting item. The Wall Street Journal reports that the EU is considering relaxing price controls on pharmaceuticals.

"In an effort to reverse the migration of Europe's pharmaceutical industry across the Atlantic. . ." is how their article starts off. And rightly so: readers will have noted my soapboxing on this point over the last few months. Despite a large and talented labor pool, and some companies with very long traditions of drug discovery, European drug research (the part that takes place in Europe, that is!) has been slipping for some years.

But I agree with the rest of the piece in wondering when (or even if) this is ever going to be implemented. This is just a proposal, and various health ministers will air things out a bit more in Rome next week. So we've started the process to begin discussions about whether to go ahead with starting the mechanism to begin drafting a law. . .well, you know the rest. This could take a while.

And even if they do ever get around to agreeing on something, it'll be worth seeing if the various EU countries actually follow up. That seems to be a real problem, various countries signing on to EU directives but then doing what they felt like doing all along (for instance, France on biotechnology.) It reminds me of a line from an old Fritz Leiber story, where some laws had a line tacked on to them during the legislative process. They all finished up with "but only if you really feel like it."

And. . .even if they get that far, how much good will it do? The governments are the big buyers, and they'll be fighting for the lowest prices they can get, naturally. The lowest price becomes the European price, since it's so easy to import and export the drugs. So with only one big purchaser per country, and an instant leveling out of the price between countries, it's hard to see how pricing freedom will make much practical difference.

In spite of all the "even ifs," I find this a significant piece of news. Even if it doesn't happen, even if governments ignore it, even if everything goes according to plan and it still doesn't do any flipping good at all. The EU (or at least part of its regulatory apparatus) has recognized three important things: that their drug industry is falling behind, that it's migrating to the US to try to catch up, and that drug prices have something to do with those first two points. (I know, these things seem as if they should be selbstverstaendlich, d'accord? But there are many people - especially in the EU, I should think - who don't see it that way.)

This won't be a politically popular position. It isn't, even here. The people that proposed it deserve some respect, even if their proposals go nowhere, de facto or de jure. But I'll be looking at another EU regulation this week that's a bit more like what I've come to expect. . .

posted at 9:11 pm

Thursday, July 3, 2003


Last year, we took our small children to Philadelphia around this time of the year. I still get questions from my 4 1/2 year old boy about "that letter that they sent to the king."

"No," I tell him, "the king didn't like it. He didn't think that the states should go off and be their own country. He sent an army, and they fought a lot of battles." (Battles are a subject of great interest to him.) "The king's army lost, and we started a new country." He and his sister now know a bit about George Washington, bits about Jefferson, about Franklin, about "that guy who signed his name really big so that the king could read it without his glasses" and some others.

They're too young for some of the things I'd like to tell them. I've already mentioned that we've never had a king since then, and that we're not going to have one, either. That's a start. And they've already grown up just knowing that if we need to, we can go where we want to, when we want. And knowing that people can grow up to do basically any kind of job that they're able to do. Small things add up: watching their parents shop for a good price on something, or watching us spiritedly hang up on a telemarketer. Watching me go to work as I tell them about trying out some new chemistry reaction. These and hundreds of other tiny examples of individual choice make their mark.

But it's hard to explain to them that so much of the world lives in conditions where people can't do such things, in the stagnant ponds that develop when these small tasks can't happen. The dead weights of poverty or of their own government (or, in many cases, of both at once) are things that my children haven't felt, and I hope that they never do. I want them to grow up the way I did: able to choose what they want to do, and where they want to try to do it. Where, if you want them, jobs are available whose whole point is doing things that no one has ever quite done before. And where you can be rewarded for succeeding, and work back to another chance if you fail.

No kings? I was wrong: we have more than two hundred million kings, all living in the same country. Happy 4th of July!

posted at 10:40 pm

Wednesday, July 2, 2003

Fluoroquinolones Are Your Friends

Well, they are when you've got pneumonia. Reading about Steven den Beste's recent struggles with it prompts me to talk a bit about the antibiotics that he's been prescribed - mainly because they've paid the salaries of a lot of people in the industry over the years! The quinolones are real workhorses for that sort of infection; he's getting the right stuff.

The most famous one is probably Cipro (ciprofloxacin,) which has been on the market for a long time now (and will be going generic soon.) But the compound class goes back a lot farther than that. Back in the early 1960s, a group at Sterling-Winthrop discovered that a compound called naldixic acid was an antibacterial. It was a byproduct of the synthesis of the malarial drug chloroquine, and it was a structural class that no one had messed with very much. In 1967 it reached the market, but it didn't exactly jolt the world of medicine.

That's because it just didn't work all that well. It had a fairly narrow spectrum of activity (gram-negative bacteria, mostly) and it was used for urinary tract infections because it didn't have very good blood levels. Even there, resistance started showing up pretty quickly. There were some variations on the structure tried during the early 1970s, but they were also aimed at the UTI market. Then in 1976, a fluorine (the med-chemist's favorite elemental substitution) was added, and that gave a broader range of activity. Two years later, an amine side chain (piperazine, a six-membered ring with two nitrogens across from each other) was added on the position next to the fluorine, and things started to get interesting. In many ways, every quinolone since then has built on that 1978 discovery. The molecule was norfloxacin. It showed the best activity yet, with some real half-life in the blood, and the race was on.

The late 1970s and the 1980s were a gold-rush era for the quinolones. Cipro was discovered in the early part of this period, and after a vicious patent fight with SmithKline, Bayer took it to market a few years later. It hits all sorts of gram-negative bacteria, including organisms that laughed at the earlier quinolones. As is the case with new antibiotics, it mowed everything down for a few years, then resistance began to pop up. Many other companies had large teams working in the area, ready to step in with new structures of their own.

But the 1990s were a rocky time for the field. For one thing, it was getting harder and harder to come up with structures that you could own. The patent situation was getting messy, with broad swaths of stuff already covered. A lot of ingenuity was applied to that problem - for example, if you want to find amines that look and act just like piperazine, but have different structures, just go to the quinolone literature. People have come up with all sorts of contorted ways to get those two nitrogens into the right place, and each novel series that worked carved out new patent territory.

The other problem was the some of the compounds developed in this period showed unexpected side effects, and were pulled from development (or, worse for everyone concerned, pulled after they'd already made it to market.) The pharmacological properties of the quinolone structures are a big squirrelly. It was already known that some of the earlier drug candidates showed phototoxicity, for example - anyone taking them had to stay out of the sun. But the problems of the new structures were worse, and they were all over the map. Serious liver problems in some patients topedoed Pfizer's trovafloxacin (Trovan,) for example, while grepafloxacin (Glaxo's Rexar) showed a rare (but alarming) correlation with sudden cardiac failure. Some new ones have made it through in the last few years, though - Bayer's "son of Cipro" (moxifloxacin, Avelox) and gatifloxacin (Tequin from Bristol-Meyers Squibb.). The newer compounds have even broader activity (hitting many gram-positive organisms as well as hammering the gram-negatives,) show higher blood levels and so on.

So how do all these things work? The targets for all the quinolones are two enzymes, DNA gyrase and topoisomerase IV. Gyrase was discovered first, back in the early days. As its name implies, it pulls and loops the bacterial DNA, which is important during replication. Otherwise, the strands get too kinked up as all the transcription/replication enzymes move along them. It mechanism is fairly audacious, involving breaking the DNA and reforming it after passing it around, and it's right in the middle of that process that the quinolones interfere. They bind, alter the conformation of the gyrase enzyme, and leave the bacterium stuck with dead enzyme and broken DNA that never got re-annealed. To make matters worse, chunks of free DNA start falling off the complex after a while; those are probably what finish the bacterium off rather than just stopping it where it was.

Topoisomerase IV was discovered a little over ten years ago. It's a similar enzyme, but it binds to DNA crossovers that form during the replication process and untangles them. Once again, the quinolones jam up the works halfway through the process, leaving another fine mess. (There are human topoisomerases, by the way, which are interesting cancer targets. The quinolone drugs don't hit these to any useful degree, though - you have to look elsewhere for really active structures.)

Bacteria have found all sorts of ways around having themselves destroyed by these mechanisms. Bacteria, I'm convinced, can find their way around damn near anything (all the way up to scorching blasts of gamma rays - see D. radiodurans.) Mutations of certain amino acids in the gyrase and topo IV enzymes are a common way out. In these forms, the binding of the quinolones is reduced, or isn't enough to cause the enzyme's action to shut down. Another common evasion is the efflux pump - proteins I've mentioned before on the site which actively shovel drug molecules back out of the cell. Some of these are particularly efficient at clearing out quinolones.

Drug research has slowed in the quinolone area, due to a general sense that a lot of the good structures (scroll down in that link) have already been tried or taken. Some groups are still whacking away at them, though. No one can say where the next similar breakthrough might come from (or how long it'll take - look how long the quinolone saga took.) But if someone finds a way to get a handle on those efflux pumps, that'd be a good candidate. . .

posted at 9:44 pm

At Least They Aren't Toxic

Following up briefly on yesterday's post, I should add that I use plenty of reagents that I don't have to worry about in my final products. These things will eat a hole in my shirt, change the color of my hair, or burst into flame spontaneously if I drop them on the floor. But they're not what you'd call toxic.

It's the difference between reactivity and toxicity. Sodium hydroxide isn't intrinsically toxic - after all, it's just the sodium salt of water. But it can be just as corrosive as anything you'd want: it'll take the outer layer of your skin right off (which is why base solutions like that feel so slippery,) and if you leave it there, it'll work its way on down to the lower layers, too. Not recommended. But it's not a compound we generally worry about for drug purity, because it washes right out. And if tiny bits of sodium counterion hang around, who cares? (That's as opposed to lithium, as I mentioned yesterday.)

So we use the whole lively arsenal of organic chemistry reagents - organometallics that leave flames burning on the end of the syringe, syrupy acids that'll cut through your lab coat like tailor's shears. It's not like we medicinal chemists shy away from that stuff - that stuff is the toolbox of organic synthesis. It would be like not plugging in your power tools.

We just have to watch out for the really poisonous ones. Actually, we just have to watch out for the persistent poisonous ones. I don't get an extra bounce in my step when I use phosgene, for example. But I don't worry about it contaminating my final drug candidate, because it's too reactive and volatile to hang around and do that. I just worry about it getting out of my hood and killing everyone, which admittedly is something to worry about. But that's why we have hoods, and pre-made solutions of the stuff so we're not handing the free gas, and phosgene detectors to let us know if anything's escaping, and so on. It's a problem with well-worked-out solutions. But having one percent thallium, say, in your drug sample is a problem whose only solutions are to clean the stuff up some more or to throw it away.

posted at 8:30 am

Tuesday, July 1, 2003

None For Me, Thanks

Looking through some lists of reagents, I got to thinking about the ones that I just won't order, and just won't use. The list is longer than it used to be when I was a graduate student, but not because I've necessarily gotten any more cautious (or any smarter.) No, when you're making potential drug candidates, you have to look out for using really toxic stuff, because you have to make sure that none of it appears in your final product.

That's really a concern with toxic elements. Toxic compounds can break down and end up as something relatively harmless, but there's no way to fix lead, for example. All lead compounds are bad; it's just a matter of degree. The same goes for mercury, which is why such arguments have raged over the vaccine preservative thimerosal. There's no doubt that it's toxic at some dose, since it's an organomercury compound: the question is, what dose might that be? (I spent some time on this issue on my Lagniappe site late last year, and I'll likely return to it. I still owe some information to Dwight Meredith over at PLA, for example.)

We'd really have to make a case for it if we want to use mercury. It had better be essential, and it had better be at the beginning of the synthesis where it has a chance to get cleaned out. I really can't imagine a true production-scale process using a mercury reagent, though, so it's better not to even start off with one on the small scale. That just delays the day when you'll have to get rid of it. Mercury reagents do some really neat chemistry, but I'm not going to experience any of it.

So none of that. And no beryllium, no tellurium, and no thallium either. Thallium does some really slick transformations on aromatic rings, but they're on the other side of the glass as far as I'm concerned. Selenium gets used, for neat reactions that's only it can do really well, but the process folks hate you when you do it. They're surely going to have to ditch it on scale. Selenium is an odd case: it's an essential trace element in the diet, and it's also poisonous. Explaining that situation to someone is a good way to introduce them to the complexities of real biochemistry.

Even our beloved palladium and platinum catalysts get the fishy glare when they're used too late in a synthesis. These really do get used on large scale in drug production, but the purity standards are set pretty tight for the metals in the final product. It helps that they're used catalytically, of course, but even then the stuff can hang around. (I can imagine the feedback I'd get if I proposed scaling up a stoichiometric platinum reaction. First I'd get run over for the expense, and then they'd back the truck up and hit me again for the purification. No thanks.) For pharmacological reasons, you don't want to use lithium in your last step, either, since it has effects all its own.

The upshot is that I haven't added nearly as many elements to my "life list" as I would have if I'd stayed in academia. I'm sure that I'd have used some of the weird ones by now, just for the experience. But it's not to be. For me, it's sulfur, oxygen, nitrogen, phosphorus. And good old carbon and hydrogen. That's where I spend most of my time.

posted at 7:40 am

Copyright 2003-2004 Derek Lowe. All rights reserved. Terms of use

Archives, RSS & Email

Click here for access to our archives.

Enter your email address below to subscribe to In the pipeline!

powered by Bloglet

Subscribe to In the Pipeline - RSS feed

Medicine & Pharma

Charles Murtaugh
The Bloviator
Zack Lynch
Richard Gayle
Randall Parker
Corey Nahman
Black Triangle
Carl Zimmer
Matthew Holt
My Lab is on Fire

Science & Engineering

Sci-Tech Daily
Jay Manifold
Paul Orwin
Rand Simberg
J. Bowen
Chad Orzel
Beauty of Gray
Uli Iserloh

Current Events

WSJ Best of the Web
Andrew Sullivan
Mickey Kaus
Matt Welch
Virginia Postrel
Tim Blair
Max Power
Asymmetic Info
Steven Den Beste
N.Z. Bear
A Dog's Life
Colby Cosh

News Sources

American Scientist
BBC - Sci/Tech
Bio-IT World
Boston Globe
Btech Investor
CBS MarketWatch
Corante Biotech
ILF News
Mercury News
Moreover - Biotech
MSNBC - Tech
Nature: Biotech
NY Times - Science
Red Herring
San Fran. Chronicle
The Scientist
Science Magazine
Scientific American
Signals Magazine
The T Sector
Wall StreetJournal
Wash. Post - Biotech


About: Biotech/med
Ag Biotech InfoNet
Amer. Journ. of Bioethics
Bioresearch Online
Biotech Chronicles
Biotech Dictionary
Biotech Journals
Biotech Knowl. Cen.
Biotech Progress
Biotech Resource
BioTrack.org - OECD
BioView - Jobs
Btech Investor
Council for Biotech
Elec Journal of Biotech
Fool.com - Biotech
Genetic Engin. News
National Agri. Library
Nature Mag - Biotech
Signals Magazine
TECSOC - Biotech
TIGR Database
WWW Virtual Library


Austral. Biotech Assoc.
Biotech Industry Org.
Cold Spring Harbor
Human Genome Proj.
Nat. Bioethics Advis. Comm.
Sanger Centre
Scripps Research Inst.
Soc. for Biomaterials
Weizmann Institute
Whitehead Institute

Scientific Quotations
Dictionary of Scientific Quotations

If the Universe is Teeming with Aliens. . .
If the Universe is Teeming with Aliens
(Then where is everybody?)

The Night is Large
The Night is Large
A Martin Gardner collection

Scientific Anecdotes
Eurekas and Euphorias