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

« Making Tumor Cells More Vigorous Through. . .Chemotherapy? | Main | Is Ampyra Any Good? »

August 13, 2012

Donald Light Responds on Drug Innovation and Costs

Email This Entry

Posted by Derek

Here's a response from Prof. Light to my post the other day attacking his positions on drug research. I've taken it out of that comments thread to highlight it - he no longer has to wonder if I'll let people here read what he has to say.

I'll have a response as well, but that'll most likely be up tomorrow - I actually have a very busy day ahead of me in the lab, working on a target that (as far as any of us in my group can tell) no one has ever attacked, for a disease that (as far as any of us in my group can tell) no one has ever found a therapy. And no, I am not making that up.

It's hard to respond to so many sarcastic and baiting trashings by Dr. Lowe and some of his fan club, but let me try. I wonder if Dr. Lowe allows his followers to read what I write here without cutting and editing.

First, let me clarify some of the mis-representations about the new BMJ article that claims the innovation crisis is a myth. While the pharmaceutical industry and its global network of journalists have been writing that the industry has been in real trouble because innovation has been dropping, all those articles and figures are based on the decline of new molecules approved since a sharp spike. FDA figures make it clear that the so-called crisis has been simply a return to the long-term average. In fact, in recent years, companies have been getting above-average approvals for new molecules. Is there any reasonably argument with these FDA figures? I see none from Dr. Lowe or in the 15 pages of comments.

Second, the reported costs of R&D have been rising sharply, and we do not go into these; but here are a couple of points. We note that the big picture, total additional investments in R&D (which are self-reported from closely held figures) over the past 15 years were matched by six times greater increase in revenues. We can all guess various reasons why, but surely a 6-fold return is not a crisis or "unsustainable." In fact, it's evidence that companies know what they are doing.

Another point from international observers is that the costs of clinical trials in the U.S. are much higher than in equally affluent countries and much higher than they need to be, because everyone seems to make money the higher they are in the U.S. market. I have not looked into this but I think it would be interesting to see in what ways costly clinical trials are a boon for several of the stakeholders.

Third, regarding that infamously low cost of R&D that Dr. Lowe and readers like to slam, consider this: The low estimate is based on the same costs of R&D reported by companies (which are self-reported from closely held figures) to their leading policy research center as were used to estimate the average cost is $1.3 bn (and soon to be raised again). Doesn't that make you curious enough to want to find out how we show what inflators were used to ramp the reported costs up, which use to do the same in reverse? Would it be unfair to ask you to actually read how we took this inflationary estimate apart? Or is it easier just to say our estimate is "idiotic" and "absurd"? How about reading the whole argument at and then discuss its merits?

Our estimate is for net, median corporate cost of D(evelopment) for that same of drugs from the 1990s that the health economists supported by the industry used to ramp up the high estimate. Net, because taxpayer subsidies which the industry has fought hard to expand pay for about 44% of gross R&D costs. Median, because a few costly cases which are always featured raise the average artificially. Corporate, because a lot of R(eseach) and some D is paid for by others "“ governments, foundations, institutes. We don't include an estimate for R(eseach) because no one knows what it is and it varies so much from a chance discovery that costs almost nothing to years and decades of research, failures, dead ends, new angles, before finally an effective drug is discovered.

So it's an unknown and highly variable R plus more knowable estimate of net, median, corporate costs. Even then, companies never so show their books, and they never compare their costs of R&D to revenues and profits. They just keep telling us their unverifiable costs of R&D are astronomical.

We make clear that neither we nor anyone else knows either the average gross cost or the net, median costs of R&D because major companies have made sure we cannot. Further, the "average cost of R&D" estimate began in 1976 as a lobbying strategy to come up with an artificial number that could be used to wow Congressmen. It's worked wonderfully, mythic as it may be.

Current layoffs need to be considered (as do most things) from a 10-year perspective. A lot industry observers have commented on companies being "bloated" and adding too many hires. Besides trimming back to earlier numbers, the big companies increasingly realize (it has taken them years) that it's smarter to let thousands of biotechs and research teams try to find good new drugs, rather than doing it in-house. To regard those layoffs as an abandonment of research misconstrues the corporate strategies.

Fourth, we never use "me-too." We speak of minor variations, and we say it's clinically valuable to have 3-4 in a given therapeutic class, but marginal gains fall quite low after that.

Fifth, our main point about innovation is that current criteria for approval and incentives strongly reward companies doing exactly what they are doing, developing scores of minor variations to fill their sales lines and market for good profits. We don't see any conspiracy here, only rational economic behavior by smart businessmen.

But while all new drug products are better than placebo or not too worse than a comparator, often against surrogate end points, most of those prove to be little better than last year's "better" drugs, or the years before"¦ You can read detailed assessments by independent teams at several sites. Of course companies are delighted when new drugs are really better against clinical outcomes; but meantime we cite evidence that 80 percent of additional pharmaceutical costs go to buying newly patented minor variations. The rewards to do anything to get another cancer drug approved are so great that independent reviewers find few of them help patients much, and the area is corrupted by conflict-of-interest marketing.

So we conclude there is a "hidden business model" behind the much touted business model, to spend billions on R&D to discover breakthrough drugs that greatly improve health and works fine until the "patent cliff" sends the company crashing to the canyon floor. The heroic tale is true to some extent and sometimes; but the hidden business model is to develop minor variations and make solid profits from them. That sounds like rational economic behavior to me.
The trouble is, all these drugs are under-tested for risks of harm, and all drugs are toxic to one degree or another. My book, The Risks of Prescription Drugs, assembles evidence that there is an epidemic of harmful side effects, largely from hundreds of drugs with few or no advantages to offset their risks of harm.

Is that what we want? My neighbors want clinically better drugs. They think the FDA approves clinically better drugs and don't realize that's far from the case. Most folks think "innovation" means clinically superior, but it doesn't. Most new molecules do not prove to be clinically superior. The term "innovation" is used vaguely to signal better drugs for patients; but while many new drugs are technically innovative, they do not help patients much. The false rhetoric of "innovative" and "innovation" needs to be replaced by what we want and mean: "clinically superior drugs."

If we want clinically better drugs, why don't we ask for them and pay according to added value "“ no more if no better and a lot more if substantially better? Instead, standards for testing effectiveness and risk of harms is being lowered, and "“ guess what "“ that will reward still more minor variations by rational economic executives, not more truly superior "innovative" drugs.

I hope you find some of these points worthwhile and interesting. I'm trying to reply to 20 single-space pages of largely inaccurate criticism, often with no reasoned explanation for a given slur or dismissal. I hope we can do better than that. I thought the comments by Matt #27 and John Wayne #45 were particularly interesting.

Donald W. Light

Comments (72) + TrackBacks (0) | Category: "Me Too" Drugs | Drug Development | Drug Prices


1. TJMC on August 13, 2012 7:26 AM writes...

Wow. I read through once, trying to parse out the meaning of any number of sentences above. Maybe I need another cup of coffee, but does anyone else find them "somewhat" unclear or overly complex for arguing a point?

Permalink to Comment

2. John Tucker on August 13, 2012 7:31 AM writes...

David, there is a simple and straighforward way to estimate the price of developing a drug. Look at the retained earnings (accumulated losses) of biotechnology companies bringing their first drug to market. This approach has several advantages.

1) As precommercial entities, such companies have little or no costs associated with any function other than drug development.

2) It vastly simplifies allocation of expenditures.

2) This obviates the issue of whether research and R&D spending are being commingled, as if you have no products, you aren't likely to be doing significant marketing.

Looking at the data, you will find that few companies bringing their first pharmaceutical product to market have accumulated deficits of less than $500M. Adding in those conpanies that have spend many tens or hundreds of millions of dollars failing to bring even a single drug to market, the $1B number begins to look low.

Secondly, if drug discovery is a free money machine, simple economics theory says that other companies would be moving into the area to take advantage of the bonanza. Instead, those oompanies already in the area are consolidating and cost cutting.

Lastly, what is this "international pharma journalist network" you are referring to? Do you really think this industry is getting glowing press coverage? Have you picked up a copy of the NYTimes lately?

Permalink to Comment

3. John on August 13, 2012 7:39 AM writes...

Sorry, that should read "whether research and marketing expenses are being commmingled".

Also should add that these financial statements are independently audited by firms that are liable to shareholders for any material mis-statement.

Permalink to Comment

4. Nick K on August 13, 2012 7:59 AM writes...

#1 TJMC: No, it's not you. I couldn't make sense of it either, but then I'm not clever enough to be a Professor of Bioethics. And Light still hasn't answered my question.

Permalink to Comment

5. overthetop on August 13, 2012 8:09 AM writes...


Permalink to Comment

6. Rick Wobbe on August 13, 2012 8:23 AM writes...

As too few others have already said, there's tremendous value in the debate unfolding here and it's not served by insults, so let's try to keep this going on a civil level. I respect Light's eagerness to engage this forum, even if I disagree (in some cases strongly) with some of his sources and assumptions. In my view, there are several fundamental, deviously tricky items that need immense clarification before we can raise the discussion to a respectable intellectual level:

1. Which estimates of R&D costs are the "right ones"? Absolutely nowhere is it possible to find auditable (i.e. verifiable), consistent cost figures for this. Tufts and PhRMA use self-reported numbers from a handful of companies, each selected according to the company's own budgeting idiosyncrasies, whereas Light uses figures and assumptions from a range of diverse, but equally idiosyncratic, sources.

2. Which tallies of new/novel/not-me-too drugs released per year are "correct"? Again, the figures we've seen are the result of very subjective analyses that can vary by multiples, not percentages.

3. What influence do arcane (to most of this blog's readers anyway) and arguable assumptions play in the financial calculations that yield the "costs" used in this argument? At least two - the price index (inflator) and opportunity costs - can be framed wildly differently by well-informed, objective people to yield figures that vary by at least 2-3 fold.

Taken together, these first three items can yield cost estimates that vary by AT LEAST the range between the PhRMA and Light estimates of the cost to discover and develop one drug. If you want to rationally and objectively identify and solve The Problem, or even IF it exists, you've got to have better data than the piles of dung we're currently flinging around.

4. Finally, it's worth taking a deep breath and asking if pharma management might done things that look venal and short-sighted without intending to, perhaps based on flawed, but well-intentioned, business strategies? On this, I think Light's comment, "We don't see any conspiracy here, only rational economic behavior by smart businessmen.", is worth unpacking. Joe Nocera had a nice piece in Sunday's NY Times on the "maximize shareholder return" mantra that bears directly on this.

Without more solid answers to the first three questions and prejudiced answers to the fourth, the argument is pitifully "full of sound and fury, signifying nothing".

Permalink to Comment

7. Anonymous on August 13, 2012 8:26 AM writes...

My question is, If it is so darn cheap and easy to develop drugs in novel therapeutic areas, why does Prof. Light not start a company to cure all of the unmet medical needs? For only $43 million, he should be able to cure cancer, dementia, or AIDS, correct?

For all of the people I hear that criticize the pharma industry, none of them I am aware of has EVER brought a drug to the clinic. Even the non-profit groups like Gates, MJ Fox's PD group, Welcome Trust, ADA, and several others groups have all spent far more than that to discover drugs for specific areas, and I don't see many small molecules coming from that. (They do have some vaccines in areas, but those are very different programs than small molecules.)

I'm the first person to admit that the pharma industry has it flaws, mostly related to the higher up management of them by boards and CEOs. But I have also seen the governments way of doing research and would like to see how many drugs the NIH and other government groups have produced over the years for their spend. I have seen first hand the inefficiencies of those programs- compared to the government, big pharma is a model of efficiency and cost savings.

Permalink to Comment

8. RB Woodweird on August 13, 2012 8:45 AM writes...

Dr. Light,

I am not an expert on drug costs - there are enough habitual readers of this blog who are that I will not set foot into the finances of the argument - but unfortunately I have become by necessity an expert on obfuscation and detecting (what I will call by its trivial name with the express disclaimer that I am not characterizing your particular argument) bullshit.

So let me apply that background to some parts of your post which jingled the old spidey sense.

"It's hard to respond to so many sarcastic and baiting trashings by Dr. Lowe and some of his fan club, but let me try. I wonder if Dr. Lowe allows his followers to read what I write here without cutting and editing."

Now that's just the kind of beginning a neophyte poster to almost any Internet forum uses to try and soften up the room. It is a derivative of the classic "poisoning the well" fallacy in logic. The poster with a cogent and persuasive argument does not stoop to such tactics. One ignores the haters - because evidence trumps snark hard.

"While the pharmaceutical industry and its global network of journalists...."

Whoa. Did you just play the conspiracy card? That is a rookie move. You are replying here to an audience of hardened professionals who know that committees and group decisions usually lead nowhere and that any conspiracy larger that three falls apart due to native human incompetency.

And so on. I was going to continue in this vein, but I have to go back to work. I get paid to actually produce products people are willing to pay for. I was going to say something like you know how that is, but then I looked up your CV. Sociology? Really? A boiling knot of sarcastic prose had to be suppressed with a superhuman effort when I read that, but as I said, I will take the high road here.

So, to sum: Bring your facts, leave the spin.

Permalink to Comment

9. Rick Wobbe on August 13, 2012 8:58 AM writes...

#7, Your post illustrates the need to make sure we are arguing with the right facts. For example, small molecule R&D expenditures by non-profits...

Gates foundation: $1.48 B (2010 Annual report) on drugs, vaccines and prevention in at least 11 areas.

MJ Fox: $49 M (2010 Annual Report)

Wellcome Trust: $700 M (2010 Annual Report) on all Science activities, of which drug research is a fraction.

Please note that every one of these figures includes a large amount of non drug-related R&D spending. So according to the best primary data available, non-profits actually spend less than the pharma industry, not "far more", as you state.

I don't mean to pick on you, I just want to emphasize the need to make our points based on observable facts, not suppositions.

Permalink to Comment

10. Calvin on August 13, 2012 9:01 AM writes...

Popcorn and a extra large soda at the ready. This is going to be good!

Permalink to Comment

11. Jim on August 13, 2012 9:01 AM writes...

Dr. Light -

Ultimately, I think you want the same thing we all want. I wholeheartedly disagree with many of your points and assertions and I am happy to agree to disagree. However, I challenge you to improve your arguments.

You use exhaustive mathematics and hard figures to describe the cost of developing a drug. Which numbers you use may be up for debate, but you use disreet, quantifiable terms. Then, when you get to the hear of your argument, you make statements like this:

"Most new molecules do not prove to be clinically superior. The term "innovation" is used vaguely to signal better drugs for patients; but while many new drugs are technically innovative, they do not help patients much."

Most, better, many, much.

The interesting thing is that those terms are qualitative descriptors of traits that are absolutely measurable. I'd be very interested in seeing how you define an acceptable level of improvement in patient outcomes and what level of harm is acceptable using some sort of risk-benefit analysis.

Try the following terms in that analysis:

NNH, NNT, survival, PGIC, SF-36

Permalink to Comment

12. John Tucker on August 13, 2012 9:08 AM writes...

Rick, what issue do you have with the means of estimation that I suggested above? It contains no cost of capital, no price inflator, nothing but the audited, accumulated expenditures of companies bringing their first drug to market, having had little or no non-R&D expenses up to that time point.

It surely underestimates costs due to survival bias, but at least puts in a floor that is incompatible with Dr. Light's number.

Permalink to Comment

13. ba on August 13, 2012 9:23 AM writes...

Ummm...#9 Rick-
The statement from #7 was "Even the non-profit groups like Gates, MJ Fox's PD group, Welcome Trust, ADA, and several others groups have all spent far more than THAT..." You left out the critical "that." #7 was referring the the $43M figure to bring a drug to market - re-read the post.

If that number is believable, assuming only HALF of the money YOU cited from the named non-profits was spent on R&D, said non-profits having now spent a combined $1.11B (one half of $1.48B + $49M + $700M) on R&D should have by now discovered and developed approximately 26 new drugs. I would love to know the identity of these 26 new drugs.

Permalink to Comment

14. Curious Wavefunction on August 13, 2012 9:24 AM writes...

I would like to point out just two things which Prof. Light still seems to misrepresent or not fully appreciate:

1. The "crisis" is based not just on a declining new drugs/investment ratio but on very real technical problems faced by scientists. The fact is that it has become harder to discover new drugs, the regulatory hurdles are higher and the kind of targets that we are addressing are much more challenging than before. The crisis therefore is based very much on actual problems that scientists in the trenches are facing. Again, Light would have appreciated this had he talked to some scientists.

2. Prof. Light seems to still think that we scientists actually *like* to mint money off minor variations. Why would this be the case? If there were really a new major treatment for cancer that was different from all existing treatments we would be dropping everything and working on it, both for scientific AND financial reasons. The main reason we develop cancer drugs that are minor variations on existing ones is, again, because it's just damn hard to come up with a breakthrough cancer drug with minimum side effects. Prof. Light seems to think that the industry "rewards" these minor variations, but that's because there's little else that it can reward. If there was anything that was a non minor variation, there is no reason to think the industry wouldn't start salivating after it.

Permalink to Comment

15. Rick Wobbe on August 13, 2012 9:30 AM writes...

John Tucker, #12, It may well be a reasonable estimate, but there's not enough information for me to judge that. What's the source of the $500 M figure? (citation please) does it include public and private companies? (I'd prefer only audited figures, which means public companies) is that just R&D spending for drugs that have made it to market or does it include in-process programs? (the latter increases the numerator of the "dollars/drug" equation and shades into the hot topic of capital costs/cost of failure that bugs many people about the TCSDD figures)

Moreover, it represents spending by a subset of the industry, with it's own peculiarities, including high up-front costs, different success rates, different components to its "R&D" budget (e.g. some Pharma companies have been shown to include marketing activities in "R&D"). As such, your estimate might be a floor, a ceiling or a median against which to compare Light's numbers. I know that sounds lame, but having looked at these numbers for so long as a perverse kind of psychotherapy, it's clear to me that we're just talking past each other unless we put much more effort ensuring we compare apples to apples.

Permalink to Comment

16. David on August 13, 2012 9:58 AM writes...

Dr Light put forward a specific number for the cost of bringing a new drug to market – 43M USD. He was ridiculed for it, and his defense is a smokescreen of irrelevant obfuscation. “We don't include an estimate for R(eseach) because no one knows what it is…” We are treated to a litany of complaints, none of which help us calculate the cost of R&D: layoffs… not enough innovation … US clinical research is too expensive… not enough safety testing…

Here’s a fact: I’ve supervised budgets for single trials in phase II and III that are greater than 43M USD. So of course Dr Light’s estimate is preposterous. And his defense is even more preposterous. What he wrote was off-topic, incoherent, and useless.

(and a respectful tip of the hat to Nick in #4)

Permalink to Comment

17. LeeH on August 13, 2012 10:04 AM writes...

A couple of quick observations:

(Light) "the costs of clinical trials in the U.S. are much higher than in equally affluent countries and much higher than they need to be"

By what estimation? And we already run many trials overseas, so are we being consistently ripped off, and don't benefit by competition? That's hard to believe.

(Light) "I have not looked into this but I think it would be interesting to see in what ways costly clinical trials are a boon for several of the stakeholders."

Hmmm. He admits that he hasn't done any real research, but he has already decided that there is a conspiracy. Considering that many trials are run in much less expensive environments (such as eastern Europe), it's hard to imagine that these costs could come down that much.

(Light) "The low estimate is based on the same costs of R&D reported by companies (which are self-reported from closely held figures) to their leading policy research center as were used to estimate the average cost is $1.3 bln.... Doesn't that make you curious enough to want to find out how we show what inflators were used to ramp the reported costs up"

Again, in the absence of hard data he has already concluded that the numbers are inflated. However, he seems to ignore the fact that most big pharma companies are public, and so reasonable estimations are rather straightforward. Unless one believes that the companies are actually spending the billions (that they claim are going to RDP) on giveaways to doctors, junkets, and other frivolous things.

(Light) "But while all new drug products are better than placebo or not too worse than a comparator, often against surrogate end points, most of those prove to be little better than last year's "better" drugs, or the years before"

This is probably true, this has nothing to do with the cost of finding a drug. I tend to have left leanings, but these are businesses, and what drugs they decide to bring to market are between them and the stockholders. It may come as a shock to Dr. Light, but it is as expensive to bring a "me too" drug to market as it is to bring a completely novel one.

(Light, from his web site) "half of the industry’s average cost of R&D is not real R&D costs at all, but an estimate profits foregone – a highly inflated estimate of what companies would have made had they put their money in an index fund and not developed new drugs in the first place"

Dr. Light seems to be happily oblivious to the reality of running a business. The estimation of profit and loss always includes the cost of money.

(Light, from his web site) "Much of that (research) cost is borne by others -- NIH, other national research programs, venture capitalists funding bio techs, foundations, and others"

No, Dr. Light. Most research is NOT paid for by others. While the initial discovery of the drug target or disease state may have come from publically-funded research, most of the cost of discovering a drug occurs after that event. And venture capitalists don't "donate" money to research, they are speculators that want a ten-fold return on their money (much of which pays for the bets that don't work out).

(Light) "The trouble is, all these drugs are under-tested for risks of harm"

Me-too drugs undergo the same clinical trials and scrutiny as the first in class. And are you saying that the FDA is soft on pharma?

(Light) "and all drugs are toxic to one degree or another."

Yes, the dose does make the poison. This is true for all drugs. If your point is that unknown toxicities are found in approved drugs, yes, that's true, but they are often effects that only show up when the drug is taken in a larger population, which is impractical in clinical trials.

(Light) "If we want clinically better drugs, why don't we ask for them and pay according to added value"

How do you assess this? And account for pharmacogenetic effects? I personally have benefitted from "me too" drugs, which helped me more than the first-line, supposedly "better" drug.

I could go on and on, but I'm betting that there will be lots of postings.

Permalink to Comment

18. Tom Womack on August 13, 2012 10:07 AM writes...

'Here’s a fact: I’ve supervised budgets for single trials in phase II and III that are greater than 43M USD. So of course Dr Light’s estimate is preposterous'

No 'of course' about it - he points out the existence of individual exceptionally expensive trials, and the claimed $43M is a *median*. What's the cheapest trial you've run?

I accept that Light's number might be wrong, but you need a more reasoned argument than that ... and it's very difficult to make that argument while pharma companies don't break down the cost of every trial in publicly-available audited documents.

Permalink to Comment

19. Jim on August 13, 2012 10:27 AM writes...

@18 Tom - you are correct, but your point made me think of something else...determining the median cost of clinical trials will be heavily influenced by the large number of Phase I trials relative to Ph II, which is larger than Ph III. It would seem to me that the true median cost is the median of Ph I trials, plus the median of Ph II plus median of Ph III.

Admittedly, I'm not sure if that's how the calculation was made or not, but for some reason, I doubt it.

Permalink to Comment

20. Boghog on August 13, 2012 10:30 AM writes...

Tom Womack, #43

The claim is that the *total* cost of developing a new drug is $43 million. This would include at a bare minimum at least one phase 2 and one phase 3 trial and more typically several of each.

Permalink to Comment

21. Petros on August 13, 2012 10:39 AM writes...

John's (#2) point is a good one. The problem can be to find companies that survive for long enough.

Vertex Pharmaceuticals was established in 1989 and finally made a profit in 2011 after successfully bringing telaprevir to market (and marketing it in the US). It had previously developed amprenavir and fosamprenavir which are marketed by GSK. And in January 2012 ivacaftor was also approved.

The cumulative reported R&D spend in Vertex' 10K reports (1997-2011) is around $5.5 billion. Simple arithmetic suggests its cost in bringing each of these drugs to market is around $1.3 billion, and the figure is clearly at least one order of magnitude greater than Light's figure of $43 million.

Permalink to Comment

22. exGlaxoid on August 13, 2012 10:41 AM writes...

7, 9 and 13:

I think that the point is that a number of non-profit groups have spent well over $43 million each in drug research, and several government agencies in the US and abroad spent well over that per year, yet few if any of them have ever produced a drug for an unmet need. The simplest way to calculate the cost per drug is to take the amount spent at a company on R & D and then divide by the drugs produced. If you look at any estimates of those numbers, for a few companies, over any period of years and divide by the number of drugs produced, you will easily see that the cost per drug for R & D is well over $43 million, and has been for decades. No matter how you claim that the costs are inflated or shifted, the cost is at least $430 million, since the 90's, and likely well over $1 Billion now. Any accountant should be able to do that math.

Having worked in research, I don;t know the exact costs of many clinical projects, but even the simplest of phase 1 studies we were doing were routinely over $1 million each by the time you include the cost of GMP oods, mandated GLP preclinical studies, and FDA paperwork.

So again, exactly what companies or groups have EVER gotten a drug to the market for $43 million? I just don't see any facts backing up that number.

Permalink to Comment

23. Rick Wobbe on August 13, 2012 10:44 AM writes...

ba, #13, I take your point. I read it differently. Funny how it perversely illustrates the importance of fully understanding the data inputs...

Permalink to Comment

24. Rob on August 13, 2012 10:54 AM writes...

Bruce Booth provided a speadsheet to estimate drug development costs on his blog last year:

No-one I know got a figure anywhere near as low as $43m and if you have to do a large PIII it is easy to get close to $2bn

Permalink to Comment

25. Anon on August 13, 2012 10:56 AM writes...

(Comment removed by request of author)

Permalink to Comment

26. Respisci on August 13, 2012 11:01 AM writes...

Dear Dr. Light.

Thank you for responding to the discussion. In your 2011 Biosocieties paper you addressed the assumptions made by di Masi et al to lower the cost of drug R&D. I propose an alternative method to calculate costs. Begin by visiting your local clinical research unit at a participating hospital and inquire what they charge a pharmaceutical company on a per patient cost for a study. You have complete freedom on the disease indication (or select a few to cover a range). What does your unit charge? $10, 000 per patient? $20, 000 per patient? Now if that per patient cost ( let’s assume $20K) was for a Phase 2 study of 50 patients you would spend $1 million, but a similar cost for a single Phase 3 study of 1000 patients would have you spending $20 million of your $43 million budget. Please note, that these are costs as set by your local participating hospital avoiding any bias set by Big Pharma. At this time, you may also want to review to estimate the average number of Phase 2 and Phase 3 trials that are performed for your disease indication.

For each clinical trial you will need to include the costs of entire team to perform data monitoring, data management and analyses. Feel free to contact various CROs and obtain cost quotes from them for one of your “studies” selected above.

Next, you will need to include the cost of a contracting a manufacturing plant that follows the regulations to ensure your drug product meets your specifications (include stability studies here). There are manufacturing plants not affiliated with Big Pharma to whom you again can obtain cost quotes.

Simlarly, contact the different CROs for the costs of complete toxicology studies –short term studies, longer term studies (these will depend on your disease indications selected by you above), repro-carci etc. Be prepared when factoring in the cost of your drug that depending on the animal species and toxic dose, you may be need to manufacture quantities greater than amount needed for your human clinical trials.

I enthusiastically await your calculations as if you are able to achieve your projection of $43 Million then I want the names of those vendors.

Permalink to Comment

27. DrSnowboard on August 13, 2012 11:02 AM writes...

What's the most expensive Phase III study?
The one that shows no reasonable efficacy in the wider patient population.

How about we try finding some solutions rather than arguing the toss over who has their unknown figures in the right unknown range?

Permalink to Comment

28. simpl on August 13, 2012 11:03 AM writes...

I suspect that most of the points of disagreement arise from different life cycle viewpoints. The crisis Derek is talking about concerns wind-downs of small-molecule research departments since 2010, especially in the US. The products they will no longer develop will be less of a concern for Donald in 2025. Similarly, desiring only the best 3-4 in a therapeutic class means that each of, say, 10 companies chasing that class should develop a lead and a handful of followers. The number