Corante

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

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: derekb.lowe@gmail.com Twitter: Dereklowe

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

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


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


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


Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem


Politics / Current Events
Virginia Postrel
Instapundit
Belmont Club
Mickey Kaus


Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Terra Slightly Less Incognita | Main | Day Off (In the U.K.) »

September 15, 2011

Targets to Avoid (Or That We Wish We Had)

Email This Entry

Posted by Derek

A discussion with colleagues recently got me to wondering about this useful (albeit grim) question: what area of drug discovery over the last twenty years would you say has taken up the most resources and returned the least value? I'm thinking more of disease/therapeutic areas, but other nominations are welcome, of course.

My own candidate is the nuclear receptor field, where some of that time and effort was mine. When I think of how enthusiastic I was ten years ago, how impatient I was to get in there and start up a big effort to really understand what was going on, to dig into the details and come up with drug candidates - and then when I think of what happened to the people who actually did that, well, it's food for thought. For those outside the field, a vast amount of effort and treasure was spent trying to work out a lot of insanely complex biology, and well, not much has ever emerged. Things went toward the clinic and never got there. Things went into the clinic and never came back out. Some went all the way to the FDA and were turned down.

So that's my nominee. I ask this question not just to wallow in misery and schadenfreude, but to see if there are some trends that we can spot, so as to avoid such things the next time they come down the chute. Given the state of the industry, the last thing we need is another gigantic sinkhole of time and money, so a bit of early warning would be welcome.

Comments (49) + TrackBacks (0) | Category: Drug Industry History


COMMENTS

1. luysii on September 15, 2011 10:02 AM writes...

Well, our genome contains 48 of them. Some have no known ligands (but the list of these is decreasing). Others have no ligand to turn them on but are constitutively active. The group used to be called nuclear hormone receptors, because that is how they were initially found, but we now know that cellular metabolites bind some (think oxysterol receptors). Also there are nuclear receptor coactivators and nuclear receptor corepressors to consider, making things even more complicated.

Anything that controls how a specific set of genes is turned on or off (which they appear to do) is bound to be important as a drug target. The problem (as it always seems to be) is that we just don't know enough.l

Permalink to Comment

2. Curious Wavefunction on September 15, 2011 10:15 AM writes...

In my postdoc I worked on trying to drug an orphan nuclear receptor which was constitutively active. Interestingly this receptor turned out to have a back pocket quite far from the usual drug-binding pocket that could be occupied by a ligand. For the complexities of NR binding it's worth taking a look at John Katzennelbogen's adventures with the estrogen receptor ("Addressing the “Core Issue” in the Design of Estrogen Receptor Ligands", J Med Chem. 2011 Aug 11;54(15):5271-82)

Permalink to Comment

3. Jonny on September 15, 2011 10:25 AM writes...

Sadly, it seems as though much of the current Alzheimer's clinical efforts will be wasted.

Although I suppose they will at least confirm the wrongness of one or more hypotheses.

Permalink to Comment

4. anon on September 15, 2011 10:25 AM writes...

I nominate comp-chem-based virtual screening efforts for hit discovery. They are a very cheap way to not find any leads.

Permalink to Comment

5. Disillusioned on September 15, 2011 10:26 AM writes...

Can I suggest in silico screening of compounds? I've only ever seen this tool give frequent hitters as an output.

Permalink to Comment

6. darwin on September 15, 2011 10:31 AM writes...

Obesity targets in CNS-take your pick. Should have redirected efforts in developing Willy Wonka's everlasting gobbstoppers reformulated with Splenda and Olestra.

Permalink to Comment

7. johnnyboy on September 15, 2011 11:05 AM writes...

In terms of therapeutic area, I second obesity. Add up the millions (billions ?) that have already been spent, and divide by the number of commercially viable, non-toxic, efficacious fat pills that have been produced (ie. zero), and you get a mathematically inconceivable number for the amount of resources that have been wasted on this. And the spending continues unabated...

Permalink to Comment

8. petros on September 15, 2011 11:14 AM writes...

Obesity as a TA seems a valid call but that is many, mostly CNS, targets.

Although your comment on nuclear receptors reflects the efforts on retinoids, LXR, FXR and PPARs, remember that a nuclear receptor agonist forms half of one of the top selling drugs (Advair)and that the steroid receptors have been a major source of revenue for years.

Permalink to Comment

9. vscreen works on September 15, 2011 11:55 AM writes...

@anon, your experience is not matched by those of many others. I am aware of two projects where chemistry was initiated on the VS result instead of the HTS. More often it interesting new classes and sometimes it strikes out.

Permalink to Comment

10. cynical1 on September 15, 2011 12:07 PM writes...

MMPs and small molecule integrin antagonists....and of course obesity as mentioned. And let's be honest, there's never going to be a safe enough drug for obesity so just give it up already and buy them a treadmill. (The treadmill will be cheaper too.)

Permalink to Comment

11. johnnyboy on September 15, 2011 12:33 PM writes...

Just wanted to add that although Alzheimer's as a therapeutic target has returned very little value for the investment, this shouldn't be held as an indication that it is a hopeless target. I'd think it's pretty normal that a decent cure hasn't been found for a disease for which the cause is still pretty much unknown. Giving a drug to decrease brain amyloid plaques or NF tangles might be just the same as treating brain cancer with tylenol for the headaches (ie treating symptoms rather than etiology). As Jonny (#3) says, even failures in AD research can amount to progress.

Permalink to Comment

12. Curious Wavefunction on September 15, 2011 12:38 PM writes...

-They are a very cheap way to not find any leads.

The operative word there is "cheap". VS is valuable precisely because it's not as resource intensive as HTS so I would not put in on the list of low input/returns technologies. And it's not like HTS has been a treasure trove of discovery. Plus, VS is still very much a young approach that is still being validated and improved. In addition one should use it judiciously; it's pointless to accuse it of failure if it's used as a black box (using carelessly designed random libraries on random targets fits into that description). As with any other drug discovery technology, expert guidance and proper understanding of the underlying chemistry and biology are important for VS or any comp chem protocol to provide value.

Permalink to Comment

13. milkshake on September 15, 2011 12:40 PM writes...

From the kinase field: p38 is a target which will live in infamy.
Then there is the COX2 fiasco. MMPs were also mentioned. Caspases as drug targets. Interleukin receptors - and any other target else that potentially activates immune system.

Nature makes mockery of the target-based drug discovery and rational drug design.

Permalink to Comment

14. qetzal on September 15, 2011 12:53 PM writes...

@milkshake,

"History shows again and again
How nature points out the folly of men!"

;-)

Permalink to Comment

15. JIA on September 15, 2011 1:18 PM writes...

@milkshake #13 -- I don't think immune activating targets are necessarily a proven waste of time for the industry. Rather I would say the jury is still out. Granted not a lot of approvals in the past 20 years, but things we've learned are enabling current drug programs.

Ipilimumab (anti-CTLA4, activates T cells by blocking tumor signals that turn them off) is looking very good in metastatic melanoma and seems likely to be approved. And there are lots of agents in clinical trials (look up TLR9, CD40, other targets).

Permalink to Comment

16. JIA on September 15, 2011 1:20 PM writes...

Whoops, I'm behind my own times: ipilimumab IS approved already, as Yervoy. There you go -- approved immune stimulating drug.

Permalink to Comment

17. You're Pfizered on September 15, 2011 1:23 PM writes...

PAF antagonists. I'll third or 4th MMPs as well. I hope I never see or need to purify another hydroxamic acid...

I saw something recently that showed the percentages of success (PoS) for a variety of therapeutic areas throughout the industry. Typically 12% was considered average. Neurological Disorders and Pain were 6 and 3% respectively.

Given the current mantra in our industry about lack of productivity, ROI and PoS, I could easily see many companies exiting those areas completely, like they did antibacterials.

Permalink to Comment

18. darwin on September 15, 2011 1:30 PM writes...

milkshake-p38 good call, but Cox2 research and development was unquestionably successful. Vioxx worked quite well, and miraculously in some patient populations. The fiasco was a function of DTC marketing campaign, physicians passing it out like candy, sitting on/"misinterpreting" subtle signals in clincial data, corporate overreaction and Class Action hagfish waiting for a free meal.

Permalink to Comment

19. Petros on September 15, 2011 1:43 PM writes...

Milkshake's comments on p38 are pertinent but it's a reflection upon a number of factors, most particularly the issues of targeting kinase inhibition for a chronic indication rather than an acute one, and the issues of redundancy in signalling pathways rather than target class.

After all these were CSAIDs™ as per a SB trademark application even before the target was identified as a kinase. The original lead structure being SK&F 86002 and nearly all SB's early patents use the CSAID term.

Permalink to Comment

20. emjeff on September 15, 2011 2:10 PM writes...

I second the obesity target(s), not only for the sheer amount of $$ it has consumed (with nothing to show for it) but also because there is an enormously effective treatment already available which can cost close to nothing: exercise and eating right. The fiction that obesity is a disease that needs treatment is responsible for an enormouse waste of resources.

Permalink to Comment

21. Thankfully ex-retinoid on September 15, 2011 2:13 PM writes...

Ah the retinoids, because there is nothing like being utterly terrified of your own compounds.

Would have been less stressful keeping a pitbull in my hood.

Permalink to Comment

22. old man on September 15, 2011 2:38 PM writes...

@ qetzal,
Nice reference to the Cult.... Go go Godzilla!

Permalink to Comment

23. Cellbio on September 15, 2011 2:40 PM writes...

agree with p38.

Interleukins are more positive. p40 antibodies (IL-12/23) set the standard in psoriasis, IL-1 inhibitors work well in familial fever diseases (small indication), IL-4/13 may not be setting world on fire in asthma, but data to come I believe, IL-17 inhibitors still in the clinic. IL-2 a pretty tough therapy, but approved. IL-6 inhibitors, some issues, but Actemra making 300-500 MM/year in US, I think? Not a slam dunk, but a class that includes a blockbuster, midlevel and smaller but still marketed therapies.

Permalink to Comment

24. And D on September 15, 2011 2:44 PM writes...

All this hindsight is great, but I wonder what the clues are earlier on that you are really up against it in terms of chances of success?

BBB penetration as essential? Lack of progress with other mechanisms targeting the same disease?or premature enthusiasm on the part of management perhaps.

Permalink to Comment

25. KP on September 15, 2011 3:40 PM writes...

I would vote for antisense, siRNA, aptamers, panning technologies as a tremendous waste of time and resources.

Permalink to Comment

26. One Timer on September 15, 2011 3:44 PM writes...

Phosphatase inhibitors seem to be extremely challenging and I'd second the nuclear hormone receptors class as well.

Permalink to Comment

27. Chris Swain on September 15, 2011 4:04 PM writes...

One thing that struck me at the Cambridge meeting was that the amount of biology that had been expended on the target before chemistry was initiated. Some of the therapeutic targets mentioned in this blog have very poorly understood biology.

Permalink to Comment

28. hermitelement on September 15, 2011 5:12 PM writes...

Therapy Area: Obesity, Diabetes other than DPP4
Target Class: Phosphatase, Ion channel

Permalink to Comment

29. barry on September 15, 2011 8:21 PM writes...

Is no one willing to give up on mammalian transcription factors as drug targets? Tularik built a whole research company on that quest and found nothing after a couple of million cmpds screened. Maybe we were able to evolve as we did because the trancription step is so well guarded from small-molecule modulators?

Permalink to Comment

30. Cellbio on September 15, 2011 9:57 PM writes...

@Chris#27,

I am usually struck by how much people rely on experimental means other than compound pharmacology to try to devine the target worthiness of targets. It is clear that however useful, these methods are poor surrogates for how a compound against a target will influence human disease. It is hard enough translating from animal models to humans with compounds, but worse yet to rely on genetic ablation, reduction by siRNA or other approaches that have mechanisms different than pharmaceutical means. Then add that there is essentially no compound against a single target, only compounds whose activity at a desire target is known, with perhaps other target impacts known, and other pharmacology totally unappreciated.

Permalink to Comment

31. Olivier Boss on September 16, 2011 12:11 AM writes...

- Interesting discussion, Derek. Some comments are good (Cellbio on cpd pharmaco, luysii, milkshake on target-based drug discovery).
- Overall, CNS targets are the most challenging (also because drugs need to get into it). If, in addition, there are no predictive animal models like most neurodegenerative diseases, good luck.
- Obesity drugs tried and (mostly) failed, yes BUT CNS targets (i.e., diet drugs).
- I won't comment much on the comments re. obesity but it's striking how everyone (non-experts) claim to know how to treat obesity, bla, bla, without knowing anything about the adaptive metabolic adaptations to changes in food intake and body fat levels (both animals and humans). Also, if people knew how to lose weight and KEEP IT OFF, how come rates of obesity-diabetes keep increasing all over the world. Would be happy to discuss all aspects in other venue.
- Nuclear hormone receptors: I agree with Derek overall. And I see his point. But there have been great successes in this family: fibrates (PPARa), rosi-pio (TZD) have helped many patients (even if not perfect), glucocorticoids (GR), androgens (AR modulators), and others. Interestingly, best successes were achieved not by developing agents against known molecular target (NHR), but rather, with functional assays/models (target was not known at that time). Now we know a lot more about NHR, and appreciate the complexity (activate, inhibit or modulate?), so it is clear to me that the way to go (as starting point) is to use FUNCTIONAL/phenotypic (and disease-relevant) cellular assay instead of other, irrelevant assays used much too widely for HTS (typically reporter gene under control of Response Element, e.g. PPRE-Luc).
- Starting with smart, relevant initial screens will save huge in resources (FTEs, $) and time, and produce more successes in drug development.

Permalink to Comment

32. Chris Swain on September 16, 2011 2:06 AM writes...

@30. Cellbio

I agree that an investigation of small molecule pharmacology is often the most useful way to explore the biology and it seems there is now much greater support for pre-competitive collaboration to put small molecule tools into the public domain.

Permalink to Comment

33. processchemist on September 16, 2011 5:02 AM writes...

Having worked on some CNS projects (a dead sector in discovery, now) I remember the NK1 NK2 fiasco (and some horrible chemistry to be scaled up in tight time frames)...

Permalink to Comment

34. Gambler on September 16, 2011 5:38 AM writes...

@30 Cellbio

In the absence of good clean tool compounds, genetic modulation of a target is really the only way to interrogate targets and pathways. I agree with you in principle, but in practice, a clean pharmacological tool can be hard to come by.

Fully agree with all comments on the need to balance target based approaches with phenotypic approaches.

Permalink to Comment

35. In Vivo Veritas on September 16, 2011 8:43 AM writes...

Ah, obesity.....
If you are willing to accept that it is a disease (and I am, putting me in company with the CDC as well as the WHO: http://www.who.int/nmh/publications/ncd_profiles2011/en/index.html) then you have to wonder what the issue is: are we unable to discover or is the FDA unable to approve?

I think we CAN discover - CB1 is a good example. Rimonabant was a pretty good drug, both in terms of weight loss and diabetes outcomes. Yes, there was a very small increase in suicidality - but it was smaller than that seen with antidepressants, and even some anti-acne drugs. So why no approval? Because the FDA is comparing to the standard of care: diet & exercise. While Rim & a few other meds clearly exceed the efficacy of diet & exercise (largely due to lack of compliance with the latter), they clearly can't beat d&e in terms of safety.

This has lead to an all-out race to minimize side effects at any cost, even at the cost of efficacy, and that's why the 3 drugs at the FDA last year were such poor performers.

My take is that FDA needs to revise the guidance - make the % weight loss criterion tougher - but clearly define the acceptable side-effect burden. Otherwise obesity rates will continue to rise, and with then diabetes rates, and with them health care costs. Sure, all this keeps me in business, but at some level, I'd like to be put out of business for doing my job too well.

Permalink to Comment

36. Cellbio on September 16, 2011 9:09 AM writes...

Gambler,

I agree that those methods, and "tool" compounds are useful in interrogating pathways and targets, just don't think that interrogation is very useful; not devoid of utility, but, imo, it should be used much less, or at the minimum, be complimented with pharmacology based screening.

My thoughts, somewhat radical in a target centric world, is that we develop drugs with desired pharmacological impact, so start with screens for that. Placing the target in the middle, validating, screening etc, can work, but the period of target-in-a-tube drug discovery is not associated with greater success. Do we ever have that clean compound? Is the clean compound even the right choice? How many compounds never see biology because of biochemical counter screens and selectivity filters? I think we fool ourselves into thinking more information on one side of the equation (compound target and biochemical panel profile) is properly guiding us when the complexity of biology remains a fertile ground for research (lots of unknowns).

I do think structure, selectivity panels etc all have their place, but our tools are driving the train too often.

Permalink to Comment

37. MS on September 16, 2011 1:32 PM writes...

Kind of depressing that 3 of 4 of the targets I worked on during my brief medchem career (caspase inhibitors, PPAR-delta, and p38) are mentioned in the original post and the comments.

Permalink to Comment

38. hn on September 16, 2011 3:07 PM writes...

As an academic, I really appreciate these discussions of what haven't worked. We really need more of this in academia.

Permalink to Comment

39. Ed on September 16, 2011 4:33 PM writes...

Never mind academia, people in industry are still working p38!

Permalink to Comment

40. Cellbio on September 17, 2011 12:12 AM writes...

Hn, you need us in academia. The model must change to include pharma vets in much larger numbers. No longer can we spend billions on research that highlights the 'potentially huge potential' and trains all scientists for academic positions that few will hold. Can you help?

Permalink to Comment

41. Fred on September 17, 2011 8:12 AM writes...

@Milkshake "Then there is the COX2 fiasco. MMPs were also mentioned. Caspases as drug targets. Interleukin receptors - and any other target else that potentially activates immune system."

I normally agree with Milkshake, but he's a bit off-base here.

MMP's are great targets and available for anyone to take up again. The reason they get bad press is: 1) early inhibitors caused stiffness due to complex selectivity profile problems that took a while to solve 2) early hydroxamic acid analogs were nasty, insoluble beasts with poor PK-- this was resolved by better chemistry 3) mixed results in tumor trials. MMPi's are more suited to inflammation than cancer, IMHO, at least until we understand more about cancer. Early prejudices often persist.

COX2 inhibitors also get lots of bad and undeserved press. The drugs aren't perfect, but look at aspirin's safety profile.

Inflammation is a ripe area for more work. If you have bad arthritis now, you are probably getting toxic junk like methotrexate; or tumorogenic, expensive, injectable, highly unsatisfactory treatments like Embrel or Humira.

One thing that got inflammation off-track is excessive emphasis and expectations on kinases.

There are also some great non-traditional inflammation targets that nobody is working on, AFAIK, right now.

My vote for things to avoid: CNS, because nobody has a clue what's going on there.

Going forward, there's plenty of great targets left, but, certainly some areas suffer from very inadequate biological understanding (Alzheimer's) or ridiculously overzealous regulation (anti-obesity). Yea, nuclear receptors are way scary.

Permalink to Comment

42. researchfella on September 18, 2011 12:04 PM writes...

@31 and @35 -- thanks for your valuable comments about anti-obesity drug research, so I didn't have to make similar comments myself.

I'm wondering what the reality is about high-fructose corn syrup (HFCS) as a potential big factor in the global increase in the incidence of obesity? Is HFCS a problem or not? And if so, do we need a social/political revolution against the corn farmers/lobbyists (like there was againt the tobacco industry for lung cancer) to address this problem?

Permalink to Comment

43. Fred on September 18, 2011 12:58 PM writes...

"if so, do we need a social/political revolution against the corn farmers/lobbyists (like there was againt the tobacco industry for lung cancer) to address this problem?"

Corn subsidies do not serve the public good. They aren't too popular with Mexican farmers/tamale makers under NAFTA, either. And corn to ethanol is economic tomfoolery.

But, ADM loves subsidies, so the big-business GOP won't eliminate them; and Dems get so few votes in flyover states (and rural regions of destination states) they aren't keen to eliminate them and make matters worse.

Permalink to Comment

44. GS on September 18, 2011 11:07 PM writes...

My nominee would be PDE4 and Adhesion molecule targets for respiratory disorders like Asthma, COPD. The complex etiology of respiratory disorders have only got more highlighted using these approaches.
Another nominee would be kinase inhibitors for inflmmatory conditions like Rheumatois Arthritis and respiartory disorders (p38 MAP kinases, syk kinase etc.) Dissociating efficacy from toxicity hasemianed as an unsurmountable challenge in these targets.

Permalink to Comment

45. Jose on September 19, 2011 1:30 AM writes...

Hey Fred,

This was just released, "F is for Fat: How Obesity Threatens America's Future" -it is nothing short of staggering. The changes in 4-5 years mean something very, very strange is going on.

healthyamericans ( dot org) (slash)reports(slash)obesity2010(slash)Obesity2010Report.(peedeef)

Permalink to Comment

46. Ed on September 19, 2011 7:51 AM writes...

#41 Fred - care to share any of these "non-traditional" inflammation targets with us? I am all ears!

Permalink to Comment

47. Bob on September 19, 2011 3:41 PM writes...

Glutamate

Permalink to Comment

48. Bob on September 19, 2011 3:42 PM writes...

Glutamate

Permalink to Comment

49. anonymous on October 11, 2011 6:10 PM writes...

Any target that involves blocking protein to protein interaction with small molecules

Permalink to Comment

POST A COMMENT




Remember Me?



EMAIL THIS ENTRY TO A FRIEND

Email this entry to:

Your email address:

Message (optional):




RELATED ENTRIES
Weirdly, Tramadol Is Not a Natural Product After All
Thiola, Retrophin, Martin Shkrell, Reddit, and More
The Most Unconscionable Drug Price Hike I Have Yet Seen
Clinical Trial Fraud
Grinding Up Your Reactions
Peer Review, Up Close and Personal
Google's Calico Moves Into Reality
Reactive Groups: Still Not So Reactive