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: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
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
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
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
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

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

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
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

« Parkinson's From the Environment? | Main | Shop Up Some Gels For the Paper »

December 6, 2013

Outcomes, Expensive Outcomes

Email This Entry

Posted by Derek

Well, to go along with that recent paper on confounding cell assays, here's a column by John LaMattina on the problem of confounding clinical results. For some years now, the regulatory and development trend has been away from surrogate markers and towards outcome studies. You'd think that lowering LDL would be helpful - is it? You'd think that combining two different mechanisms to lower blood pressure would be a good thing - is it? The only way to answer the questions is by looking at a large number of patients in as close to a real-world setting as possible.

And in many cases, we're finding out that some very reasonable-sounding ideas don't, in fact, work out in practice. These aren't just findings with new or experimental drugs, either - as LaMattina shows, we're finding out things about drugs that have been on the market for years. This illustrates several important points: (1) There's a limit to what you can find out in clinical trials. (2) There is a limit to what reasonable medical hypotheses are worth. (3) We do not understand as much as we need to about human biology, in either the healthy or diseased state. (4) A drug, even when it's been approved, even when it's been on the market for years, is always an experimental medication.

LaMattina also points out just how crazily expensive the outcomes trials are that can generate the data that we really need. He's hoping that companies that spend that sort of money will emerge with a compelling enough case to be able to recoup it. I certainly hope that, too - but I'm absolutely 50/50 on whether I think it's true.

Comments (16) + TrackBacks (0) | Category: Clinical Trials | Drug Prices


1. Cellbio on December 6, 2013 10:23 AM writes...

Boy, could not agree more. From reason (2) above, I reflect on the post of last week where the value of animal models was discussed. I think the use of animal models to build medical hypotheses is very limited (not zero in all cases). I do think the answer is to generate more treatment outcome data, but yes the cost is crippling. I am shocked when I see trial budgets emerge and per patient costs rise to 10s of thousand per patient.

I place lower odds than 50/50 in aggregate with this financial burden. My experience in big companies shows the willingness to reload on a program that has hit a bump in the road is overshadowed by moving a 'blue sky' program forward that has yet to show a blemish. Better get it right the first time, with far too many variables, including patient population, dosing, combo therapy, etc. as, unfortunately, ignorance of challenges looks more attractive than demonstrated limits which leads to cycling to the next opportunity rather than persisting in the face of less than stellar outcomes in the first trial. For a start-up, there really is only one, maybe two, shots to get it right before investor good will, and cash, is gone.

Maybe disease foundations, instead of funding basic research, could defray the cost of trials? Create patient registries to aid in trial recruitment and assessment of comparative efficacy? Do any do this now?

Permalink to Comment

2. NoDrugsNoJobs on December 6, 2013 10:52 AM writes...

#1 - Good point or question on the funding. Some of these studies really need to be taken up by agencies like the NIH who, by the way, have conducted some like the Woman's Health Institute Study. They can cost billions of dollars to do properly but in the case where millions of people may be taking drugs that have been never really validated vis-a-vis any kind of powered outcome study, the value is incredible.

Right now there are many men taking testosterone replacement therapy - truth is, we don't know if androgens are good, bad, when, why, etc. Same with NSAIDs. What do we know about ibuprofen but yet millions of people quaff it on a daily basis.

Just saying, there is so much we don't know. We have lotds of drugs already approved, generic and being used by millions but yet we really don't know.

Knowledge gained is knowledge forever, investing in this type of robust clinical research could lead to valuable knowledge beyond our lifetimes....

Permalink to Comment

3. Anon on December 6, 2013 10:59 AM writes...

Here a riddle: If one was an amoral machine (such as a large pharma company) how deep down the rabbit hole of (stage 4 trials) do you go?
If you prove your drug is even better than you received approval for, do you suddenly get to charge more for it? No.
However, if you find issues you will most certainly lose money.
The only thing to gain would be a new indication, however that isn't very likely as you are looking at the population you already have approval for.

The outcome of looking is costly, as is the very act of looking.

Permalink to Comment

4. Biotechtranslated on December 6, 2013 11:47 AM writes...


Sure a drug company could charge more with positive phase 4 trials, they can just take aggressive price increases if they feel the data supports it.

The other possibility is an expanded patient population or an increase in market share, just like Pfizer did with Lipitor.

There is a TON of upside for a pharma company with positive results in a well-crafted phase 4 trial

Permalink to Comment

5. watcher on December 6, 2013 11:50 AM writes...

One time I had a strong disagreement with my primary physician. He wanted me to take Zetia and I did not want to use it, saying that it had not been shown to have a meaningful effect in an outcome study. He insisted that cholesterol lowering was cholesterol lowering, so that the statin studies demonstrated the value to lowering cholesterol. I tried to explain to him that diet systemic cholesterol did not have to be the same as metabolic pools, that human physiology was often shown to keep such pools apart. He did not agree, could not understand, got rather irritated. Even so, I did not take the drug. Later, when outcome studies with Zetia were finally reported, he had to admit that he still did not understand and would continue to use the drug. So idiotic....Merck's initial selling of the value to lower cholesterol with statins was way too overdone, even with the later knowedege that statins do much more than simply lower that one biomarker, and may work in many situations as a general antiinflammitory instead.

Outcomes, outcomes, outcomes.

Permalink to Comment

6. dearieme on December 6, 2013 2:03 PM writes...

"You'd think that lowering LDL would be helpful": or not, depending on your assessment of the intellectual calibre of the arguments you've read, and on your critical assessment of the evidence.

Permalink to Comment

7. PSU on December 6, 2013 3:49 PM writes...

"A drug, even when it's been approved, even when it's been on the market for years, is always an experimental medication."

I'm not a chemist of any sort (I got through high school physics by the grace of God), but it strikes me that this sentence should be engraved somewhere that every doctor and every patient can see it, and (again, with the grace of God) understand it.

Permalink to Comment

8. dave w on December 6, 2013 5:06 PM writes...

#5: I've heard of a theory that the actual mechanism by which statin drugs can benefit circulatory health is not "reducing cholesterol" as such but the reduction of "C-reactive protein" which is associated with inflammation.

Permalink to Comment

9. Bioorganic Chemist on December 6, 2013 8:12 PM writes...

Here is a reasonable call (and need) for more data to demonstrate safety and efficacy of approved drugs. Drugs that eventually will become generics, via research that is not funded by generic drugmakers, who function in a realm of zero risk, all benefit. How about a tax on generic drugs that is used exclusively to fund research on drugs post-approval (managed by NIH, to focus on drugs with the most study needed, including those already generic)? Surely there is some way to get generics makers to actually pay for *something* in the drug development process?

Permalink to Comment

10. Lunar landing on December 7, 2013 1:08 AM writes...

@#9. That is an interesting thought. Here is another way to think about this although it may be heresy. Can large pharma be incentivized to study their drugs further in outcome studies by allowing longer patent protection? Would be in the public interest and may provide the proper financial incentives for further investigation. Would require careful crafting of new law which could be perceived as benefiting large industry, not something I expect to be well received anytime soon. Generic companies will not do the studies you are asking, margins are too thin and time to make money is too short. In the end generics are not in the business of making or understanding drugs - not demeaning generics here it is just the reality of their business model.

Permalink to Comment

11. UK Chemist on December 7, 2013 4:03 AM writes...

I don't disagree with the sentiment, but how long and large would these trials need to be? I can't see how this is sustainable with the current period of exclusivity.

Permalink to Comment

12. Anonymous on December 7, 2013 6:49 AM writes...

@9: Great idea, I like this a lot!

Permalink to Comment

13. Eric Robert Jablow on December 8, 2013 12:13 PM writes...

Have you any comment on today's Washington Post article on Avastin versus Lucentis?

Permalink to Comment

14. matt on December 8, 2013 2:14 PM writes...

I'm not sure I agree with the thrust of LeMattina's argument.

I think very thoughtfully chosen markers of disease continue to be the appropriate target for pre-approval studies. They aren't perfect, we will be wrong sometimes on some, but they are the appropriate blend of accuracy, timeliness in our understanding process, and cost effectiveness.

Also important, those studies also generate some outcome data, which the FDA will continue to watch and require. That is, the FDA in my opinion has not been too lax in allowing "cherry-picked" biomarker surrogates to be substituted for real results. This blog has covered these efforts by the FDA, including overall survival in oncology trials and clinical measures of efficacy in Alzheimer's trials.

His example of ESRD treatments shows that studies need to be done, but does not show any problem at all with pre-approval studies. It simply is unknowable before trial whether interactions will occur or not, to benefit or detriment. It certainly isn't as obvious as 1+1, to me.

Fully powered outcomes studies are appropriate after drug approval. Since they need not interfere with the patent life of the drug, drug companies (or research bodies) ought to save considerable money on enrollment. Data collected as an academic study on outcomes, in the course of regular treatment, ought to have a very different cost than data collected pre-approval to watch safety and efficacy. I think it is appropriate to have a blend of pharmaceutical and academic/public health money funding these studies.

Digital medical records, IF they enable medical researchers access to outcomes data with reduced breathing-down-the-neck of lawyers worried about privacy lawsuits, might significantly streamline the collection of some of this data.

How cholesterol medicines work is irrelevant. The drug trials to date have been effective in showing whether or not they do work. People can argue all day about what causes benefit, but the benefit itself, which is what the drug trial attempted to measure, is undeniable.

Permalink to Comment

15. Antipodean Lurker on December 9, 2013 1:05 AM writes...

#2, #7 and #14 If all medicines are considered experimental then we don't need to get informed consent to randomise patients to treatment with approved medicines and actually collect some causal data from the millions of patients who use medicines. National health systems that have large populations on medications would be able to detect outcome differences in finite times (or rule out a difference in effects of a defined size). Denmark and Sweden and New Zealand have the data linkages to do it but maybe not the populations? Judea Pearl's work on causality points to medicine assignment being randomised being very important for the establishment of causality. Of course, the influence of doctors on prescribing would be greatly reduced....

Permalink to Comment

16. Anon on December 9, 2013 10:49 AM writes...

@4. Biotechtranslated
I am speaking to the use of the same patient population in a stage 4.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry