Skip to Content

Drugs Purely From Academia

I see that there’s an investment fund in Europe saying thatEvery year, around 30 to 40 percent of the drugs approved by the FDA (U.S. Food and Drug Administration) were actually discovered in European academic labs“. Meanwhile the European Medicines Agency says that “Academia and other public-private partnerships helped to develop about 17% of the 94 products approved between 2010 and 2012“. And Médicines Sans Frontières is disputing some Pfizer advertising in the UK, saying that “tax payers foot much of the R&D bill through the funding and hard work that universities and government-funded laboratories do in actually discovering the compounds that are turned into blockbuster drugs“. Not to be outdone, I popped up yesterday saying that you could count the pure university-discovered drugs on your fingers. We can’t all be right. So it looks like it’s time to go over the real figures again.

The most comprehensive source for these remains this paper in Nature Reviews Drug Discovery by Robert Kneller, which covers the 1998-2007 period, and which I blogged about here and here. (I’m definitely not going to update its analysis while writing this blog post this morning, but I would like to do a bit of that for a future post).Kneller did a lot of work on this paper, and one of the things he did was assign each drug to a “main discovering organization” by going back to the original patent filings and publications. Looking at his supplementary Excel files, he has 252 drugs approved during that period. 117 of them were “standard NMEs (new molecular entities)”, that is, small-molecule drugs approved through the usual FDA process. Here are the ones that are primarily assigned to a university or similar:

1998: loteprednol (University of Florida)
1998: paricalcitol (University of Wisconsin)
1999: poractant (Karolinska Institute)
1999: doxercalciferol (University of Wisconsin)
1999: aminolevulinic acid (Queens University)
2000: cetrorelix (Tulane and Inst. for Drug Discovery-Budapest)
2000: cevimeline (Israel Institute for Biological Research)
2001: dutasteride (Tulane) Update: plenty of evidence for GSK here, so I’m removing it from the count below.
2001: galantamine (Mt. Sinai / Free Univ. Berlin)
2003: emtricitabine (Emory)
2005: exenatide (Mt. Sinai / Bronx VA Hospital)
2006: decitabine (Czech Academy of Science)
2006: docetaxel (CNRS, France) Update: added this one to the list; it’s a collaboration in the paper
2007: rotigontine (Groningen)
2008: lanreotide (Tulane)

That’s 14 out of 117 “standard” small molecules that don’t have drug-company fingerprints on their original discovery. The paper also breaks out 98 NMEs that were approved via priority review. Here are the ones that have a university, etc., as their primary discoverer, with no drug company in on that part:

1998: sacrosidase (Ludwig-Maximilien Univ. and Univ. of Connecticut)
1998: thalidomide, its revival (Rockefeller U. and Harvard/Children’s Medical)
1998: valrubicin (Dana Farber)
1999: ferric gluconate (Humboldt Univ. Berlin)
1999: nitric oxide (MGH and Karolinska Inst.)
1999: zanamivir (Monash Univ.)
2000: arsenic trioxide (2nd Shanghai Medical Univ. / Harbin Med. Univ.)
2000: vertiporfen (Univ. British Columbia)
2002: oxaliplatin (Nagoya City Univ.)
2002: oxybate (Univ. of Arkansas)
2003: enfuvirtide (Duke)
2003: abarelix (Indiana)
2004: pemetrexed (Princeton)
2004: apomorphine (Harvard/MGH , Brookhaven, Cagliari Univ., Univ. College London)
2005: tipranavir (Czech Academy of Science)
2006: vorinostat (Sloan-Kettering and Columbia Univ.)
2007: sapropterin (NIH and Nagoya Univ.)

That’s 17 out of 98, although I have to say that I don’t find all of those equally exciting, by any means. Some of them are slightly repurposed versions of things that had been known for a long time (although the revival of thalidomide certainly doesn’t fall in that category – that took some real imagination and spirit). This brings up a point mentioned in the NRDD paper, that the great majority of drug sales are actually from compounds discovered by the drug companies themselves. With some definite exceptions, the compounds listed above tend to be of lower impact. That may be partly because drug companies most definitely keep sales/medical need (those two are usually coupled pretty well) in mind when starting and continuing projects, and partly because some of the academic-derived compounds come under the “low-hanging fruit” category, such as the repurposing of apomorphine as a rescue therapy during severe Parkinson’s episodes (Apokyn).

You may notice that pregabalin isn’t on there, although perhaps it should be. It was listed in the spreadsheet as “Northwestern and Warner-Lambert” for earliest credit, though, and although the compound itself did come out of the Silverman lab there, the biology of it (mechanism of action, target, etc.) came out of the industrial work, and the original hypotheses turned out to be almost completely off. So that one is sort of a special case, although there are several other drugs that were also direct collaborations between academia and industry, and I’m sure that there are arguments to be made about each of them, in one direction or the other. The ones above, though, are the ones that you can say, pretty definitively, had completely academic origins – just the way that MSF is imagining it happening, to judge from their statements about Pfizer.

Now to the points in that first paragraph. These drugs represent more fingers than I happen to have, so my statement was indeed an exaggeration. I’ll update that post with a link to this one. Somewhere around 15% of the total approvals would be a more fair estimate, at least for the time frame this paper studied. Now as for that “thirty to forty per cent” coming from European labs, well. . .I sure don’t see it in the 1998-2007 data, that’s for sure.Even if you include all the joint-credit compounds not on the above lists, you don’t come anywhere close. I’ll see what it’s been like more recently, but I have my doubts. That 17% figure from the EMA might be more accurate, but even that’s high compared to this data set. Maybe those years were particularly rich? We shall see.


61 comments on “Drugs Purely From Academia”

  1. rob says:

    The VC who made the comment corrected it on Twitter to say “discovered in European labs” – not just academia

  2. David Borhani says:

    These lists seem to miss quite a few drugs. Based on personal experience, what about these drugs from Southern Research Institute (some in collaboration with SRI International and/or Memorial Sloan Kettering)?

    pralatrexate 2009
    fludarabine 2008
    clofarabine 2004
    amifostine 1995

    (And three older drugs outside the time range: lomustine, carmustine, and dacarbazine).

    1. Derek Lowe says:

      1995, 2008 and 2009 are outside the paper’s scope, for sure (approved 1998-2007). As for clofarabine, though, that one looks like it should definitely be in there. I’ll add it to the statistics in the post – thanks!

    2. Derek Lowe says:

      Actually, I missed clofarabine – just added it!

    3. Colm says:

      Ahh, I did my PhD in a lab at Southern through UAB, always brings me a bit of a warm feeling to see that we had some real hits. The structures of those and approval dates are framed in one of the entrance hallways.

  3. Erebus says:

    Some of those compounds are head-scratchers. Particularly:

    …Not only an ancient folk-drug, but something that had been purified and used the USSR as a pharmaceutical since the 50’s? Why not just give the Soviets credit for it?

    -Aminolevulinic acid
    …How can one call such a downright ubiquitous molecule a “new molecular entity” with a straight face?

    -Ferric gluconate
    …This seems like awfully low-hanging fruit to me.

  4. Anon says:

    A little correction. Tipranavir wasn’t patented by Czech Academy of Sciences (but rather by Upjohn), although decitabine, telbivudine, tenofovir, adefovir, azacitidine were. The Excel file has it right.

  5. anon says:

    What about ziagen (primarily from Bob Vince at the University of Minnesota, approved 1998)?

    1. Derek Lowe says:

      Ziagen is listed as U. Minn and Burroughs, but in light of the legal wrangle afterwards, let’s award it to U Minn, definitely! I’ve changed the post.

  6. Anon says:

    In any case, absolute output numbers are meaningless. Even the output ratio between pharma and academia is meaningless.

    Much more meaningful, is financial EFFICIENCY, in terms of output value vs input investment. Because even if academia produce only 5% of the drugs that pharma produces, it is still way more efficient if it only spends 2% of pharma’s R&D budget on drug discovery and development.

    On that basis, I find this whole exercise completely … er … academic – and pointless, not to mention potentially misleading!

    1. ab says:

      Do you have a source the corroborates your suggestion that academia’s ‘efficiency’ is higher than industry’s? How would you measure what was spent to develop a particular drug in academia? Do you count all of the money spent over a prof’s career on dead ends? Do you count all of the money spent on research at the entire university, or perhaps the entire university system? Because it disingenuous beyond the pale to think merely in terms of the singular grant (or what have you) that was spent to discover one of those singular drugs out of academia.

      1. Anon says:

        You missed the bit where I said “IF”. I don’t know IF academia is more or less efficient than pharma, but neither do the original post and analysis above tell us, and THAT is my point.

        1. CMCguy7 says:

          I can not produce hard evidence whether “academia is more or less efficient than pharma” for discovering a drug however casual reasoning would suggest academic labs are probably substantially less efficient for that task. Largely this can be assumed from the differing nature and mission of academia being focused on training students, often lacking fuller skills, and publishing papers (to enlighten others and support grant applications). With few exceptions the depth and breadth do not exist to perform all the fundamental early activities (biology screening, synthesis, etc.) in a single group and even through potential collaborations within same or other universities timely and smooth coordination is often less established than under the umbrella of industry (albeit greatly deteriorated these days) . Not that I would make any claims about significant drug discovery efficiency in industry (beyond they suck) but both organizational models likely require a dose of Serendipity and plenty of perseverance to succeed where I might argue academics could have an edge in the former with more open free-range approaches while industry (at least used to be ) more able to commit to longer term projects/greater influx or resources to generate and progress candidates. I would rather avoid the term efficiency even when attempting relative comparisons and basically see this as Apples and Oranges discussion point that often comes down to overloud people who like Apples devaluing those that prefer Oranges.

          1. Anon says:

            Most likely you are right, however it just irked me that the focus of the initial post and discussion seemed to imply that total output is what “counts” more than efficiency, and could even be interpreted as a measure of efficiency, which is certainly not the case.

            More important is the quality of the discussion to understand the implications and limitations, so I just wanted to raise a flag in case anyone reading this thought they could draw any conclusions.

    2. Weezl says:

      Your math is too simple. What was the efficiency of these discoveries in terms of time? Most academic labs are staffed with only a few people and can easily spend decades trying to get one compound to work in any usable form. Input investment, output, and time are three measures to consider for overall efficiency. How quickly does Pharma generate a drug (I know the failure rate is high, but bear with me) versus how long does an academic lab take to generate one of the drugs presented here?

    3. Phil says:

      Output value vs input investment is completely out of the scope of this argument. Even those drugs that were initially discovered in academia would never have realized any ROI without a pharma company to fund the development. You would have to separate development costs (which are actually most of the costs) to have an apples-to-apples comparison.

      Is academia more efficient at discovering compounds that end up being drugs? Given how much time academics spend on tool compounds (or worse, PAINS), my guess is “no.”

      1. Calvin says:

        I can sort of answer that. I used to work in this space on the funding side. Of the things that got funded (less than 5% of apps got funded) the success rate (compared to industry average)was similar to industry (up to clinical development) and the cost was biotech like. So academia can do it (a small subset anyway) but they require considerable external (industry) expertise the further into development they go and typically it takes longer (but not always). So the answer is probably that academia as a whole is much much less efficient (95% of translation style projects are pretty poor) but there is a subset that is competitive part of that subset is very good indeed. And the trick is finding those people and helping them.

        1. Anon says:

          Thanks for the insight, sounds very plausible.

          1. Phil says:

            Thanks, Calvin. My anecdotal experience parallels your more extensive experience. There are a few academic drug discovery centers that seem to really know what they are doing. Often they are led by highly accomplished ex-pharma scientists (at least one of whom has commented in this thread). They also know enough to know that you can’t expect to take it all the way to market on your own.

    4. FarmaBiz says:

      Not meaningless at all, but a good, perhaps most valid, baseline. One can always do a 2nd, 3rd, etc. derivative take off the baseline, but you need a baseline, especially this one which again defeats the hideous meme of Gov/Academic creation of all drugs that Big BioPharma rips off and the other meme that all Big BioPharma does is tweak existing molecules.

      Efficiency alone is not enough since it does not distinguish between the “Low Hanging Fruit Crowd” and the Big Innovators. If Gov/Academia is 0.5x the $ of Big BioPharma and they are the “Low Hanging Fruit Crowd” that they are, perhaps they should be 0.25x to reflect that. Another meme, Gov is more efficient than the private sector, is total B.S.

  7. Thanks to all today for the spirited discussion and rapid analysis in response to my query about the quote in the Reuters article.

  8. George Demetri MD says:

    Missing some important ones! Abiraterone? Enzalutamide? Although DEVELOPED by industry, these were discovered by academics. We need to be clear that discovery and development are complementary, and the skills overlap often.

    1. Derek Lowe says:

      Abiraterone was approved in 2011, and enzalutimide in 2012, so they’re outside the scope of the NRDD paper that I took these from. They’ll be part of the updates that I hope to do, though!

    2. road says:

      FDA approved in 2011 and 2012. The list spans 1998-2007.

  9. sofosbuvir says:

    The boundary between academic and private-sector R&D can be pretty grey, especially given the growing trend in public-private partnerships. Take sofosbuvir for example, arguably the hottest compound in pharma in terms of current $$ earnings. It’s Gilead’s cash cow, courtesy of Pharmassett (PSI-7977), which was in turn founded by Raymond Schinazi, an “academic”.

    1. Calvin says:

      Not quite. The prodrug chemistry was developed under the academic tag that Ray had. But they then spun that out and the drugs were all with private money. And that was no slam dunk. Took them a few clinical tries (4 I think, 2 with Roche) before they got the right nuc to attach to that prodrug bit. So the academic bit was very early and more of the enabling kind. Don’t feel bad for the university though. Their equity in Pharmasset did well!

      1. b says:

        Not quite again. The phosphoramidate prodrug moiety was developed by Chris McGuigan at Cardiff in the early 90’s. It’s been used on a bunch of nucs since then, not just sofosbuvir. This is the basis for a lot of the patent busting arguments in the UK- that the protide tech was developed by an academic lab in the UK with gov’t $

  10. MoBio says:

    Here’s another one from academia:

    Lacosamide (INN, formerly known as erlosamide, harkeroside, SPM 927, or ADD 234037) is a medication discovered by Dr. Harold Kohn while at the University of Houston and developed by Union Chimique Belge (UCB) for the adjunctive treatment of partial-onset seizures and diabetic neuropathic pain marketed under the trade name Vimpat.

    “Lacosamide was discovered by Harold Kohn and colleagues at the University of Houston in 1996..approved 2008.”

    1. Anon says:

      Came to the comment section to add Lacosamide as well…

      Here is the original paper, Lacosamide is compound 18.

      J Med Chem. 1996 Apr 26;39(9):1907-16.
      Synthesis and anticonvulsant activities of N-Benzyl-2-acetamidopropionamide derivatives.
      Choi D1, Stables JP, Kohn H.

      It is also noteworthy that it was discovered through phenotypic screening in a rat model of epilepsy. Mind you, it was not the result of the screening of a random library but rather the end product of SAR studies conducted with this in vivo assay as the functional readout!

  11. watcher says:

    Dutasteride (Avodart) was originally made by Glaxo, with original composition of matter patent filed about 1994. Any work done by academic group(s) would have been clinical studies supported by Glaxo, and later GSK, not simply done by academic groups on their own.

  12. watcher says:

    For a complete summary on the discovery of dutasteride (Avodart) see the review by S. Frye. Concept, discovery, patent history all began and was sponsored by Glaxo, later GSK.
    Curr Top Med Chem. 2006;6(5):405-21.
    Discovery and clinical development of dutasteride, a potent dual 5alpha-reductase inhibitor.
    Frye SV

  13. MoMo says:

    Where’s all of Dennis Liotta’s compounds from Emory?

    You’ll find there’s more than you realize coming from academia. Russia and China- not so much.

  14. MN says:

    Docetaxel. Although some people from Rhone-Poulenc (now Sanofi) appear on the original patent, it was fully discovered by Pierre Potier’s group at ICSN. This group had also previously “provided” Vinorelbine to Laboratoires Pierre Fabre (not in the scope of the paper).

  15. Synthon says:

    Don’t forget Ray Shinazi & Dennis Liotta’s (Emory University, Atlanta) 3TC (Lamivudine), another nucleoside analog reverse transcriptase inhibitor for HIV & Hep B…

    When I was doing my undergrad at Emory, they were the only professors on campus driving Ferraris and Porche 911 Turbos, to be sure.

    1. Derek Lowe says:

      Definitely a good example, but it was approved in 1995, outside the timeline of the paper.

    2. DrZZ says:

      They were driving those cars because Emory made a ~half billion dollar deal with pharma. I always ask the “pharma rips academics off” advocates to explain that. Was half a billion too low? I thought it only took about 50 million to find a drug. Did pharma get ripped off by the academics?

  16. Insilicoconsulting says:

    So doesn’t background biology work done at academic labs that directly or indirectly helps pharma companies bring some drugs to market count?

    While pharma certainly does a lot of target ID, validation etc a lot of new biology may still be done in academia. And if you don’t have basic biology elucidated (alzheimer’s), discovering drugs is gonna be difficult. right?

    Derek and others have certainly made a compelling case that academia historically has been weak in medchem. So these numbers ~30? are not so bad seen in that light.

  17. Ben says:

    How about darunavir/Presizta? I reckon between all the people reading this site we can get a more accurate list than any of the above references in a couple of days…

  18. A Nonny Mouse says:

    Valacyclovir (de Clercq) but slightly before the scope of the paper.

  19. JB says:

    Is Ray Schinazi emtricitabine and sofusbuvir too? Is that all the same guy? What’s he driving nowadays if so?

    1. jbosch says:

      Public transportation, as all smart people do 🙂

      This is a very interesting thread, and I am looking forward to David’s update on the newer ones.
      I think a partnership between academics and industry is the way to go to be more productive. It’s more a complementary approach. There’s hardly an academic lab with an army of people that could do what pharma does, but on the other hand perseverance on a pathway/protein target is something that can be done very well in academia at the early stages of target discovery or even mechanism of action..

  20. petros says:

    temozolomide 1999 University of Aston

    Malcolm Stevens did a pretty good med chem job on this before it was licensing to Schering Plough

  21. Georg W. says:

    I thinkTecfidera was discovered at a university hospital in Germany and then sold to Biogen (for pennies…). The active compound, dimethylfumarate, is used against psoriasis and helped also against MS in someone who suffered from both diseases…

  22. In Vivo Veriats says:

    Out of the time scope of the paper, but Mayalept? Rockefeller – leptin –> Amgen–> Amylin – metreleptin (small structural change) –> BMS/AZ – Mayalept.

  23. Stephen Frye says:

    As a co-inventor of dutasteride, I can say that GSK was the single source. Once it exceeded $1B in sales, I had the unfortunate opportunity to help prove this in court – part of the generic business model that is pretty painful = invest in lawyers, not R&D. I never noticed it listed as from Tulane in Kneller’s review. Seems like a plain old mistake since I have also never heard of anyone at Tulane contending the issue. Hard to keep mistakes out of big data sets though – maybe Google can invent a “bot” to do this?

  24. partial agonist says:

    Fusilev, or l-leucovorin, was approved in 2008. It is sold by Spectrum, but IIRC the foundational patents came from Scotland (U. Strathclyde)

    Many drugs have mixed origins of course, but however you define it, the 15-20% range sounds like a perfectly reasonable estimate

  25. ab says:

    So now that we’ve established that statements suggesting that most drugs are discovered in academic labs is complete and utter bunk, perhaps we can think further about the question, which has not been asked but should be, of whether the drugs that ARE discovered in academic labs and bought by Big Evil Pharma end up being a good deal for those academic labs. My sense is they are a very good deal for those academic labs. If they choose to license their intellectual property, they stand to make a very large amount of money if a compound makes it to market. What do people think, that academics are a bunch of idiots and are coerced into giving up their intellectual property for pennies because they don’t know the value of their discoveries? Kind of a harsh view of academics, isn’t it?

    1. Phil says:

      Silverman Hall at Northwestern is a very nice building.

      1. watcher says:

        Yes, it is nice building. But the conundrum is that a post-doc in his lab at Northwestern made the compound as an inhibitor of glutamate decarboxylase, which is not the mechanism (thought to be a channel blocker) by which the compound functions for the indications it has been approved.

        1. Phil says:

          That’s beside the point. Silverman and Northwestern
          made out pretty nicely from the deal. Hardly the victims of Big Evil Pharma.

          1. jbosch says:

            He gave a fantastic Keynote at AIMECS in Jeju past October. I was very impressed by his accomplishments and he seemed very down to earth despite the $$$

  26. WestCoast says:

    What about imatinib (Gleevec)? Developed at Oregon Health and Science University and approved in 2001.

    1. OldLabRat says:

      The earliest patent application was filed in 1992 in Switzerland and assigned to Ciba-Giegy, a forerunner of Novartis. I believe Druker at OSHU led the initial clinical trials for CML, so was involved in the development, but imatinib didn’t originate at OSHU. It’s fairly common for clinical trials to be conducted at teaching hospitals/universities.

      1. Stephen Frye says:

        Exactly! It is very common (see the “Emperor of all maladies”) for the press and laypeople to think of the “discovery” as the testing of the compound in the clinic, not the actual design and synthesis of the compound. While clinicians involved in development make huge contributions to drug development, the discovery is usually done by earlier stage chemists + biochemists + pharmacologists. I think that bit is just less familiar ground and everyone likes the idea of a clinician watching a drug being infused into a patient and shouting, “eureka!!, I discovered the cure!”. In a way, they did, but not the way we chemists think about it.

  27. Andre says:

    Derek, sacrosidase is an enzyme, not a small organic molecule. Where you intending to also include NMEs in the area of enzyme replacement therapy in the list?

  28. Roger says:

    It’s always amusing to see the arguments about who actually discovers or develops drugs – academics or (evil price gouging) pharmaceutical companies. An interesting question that has been alluded to but not addressed directly: of the (slightly more than a) handful of drugs whose origins are attributed to academia, what was the level of investment that academia dedicated to discovery, IND filing, CMC and clinical development and NDA filing? For the same drugs, what was the split of total investment from academic sources vs pharmaceutical sources (partners/licensors)?

  29. Jeff says:

    As for Pregabalin (Lyrica for the non pharmacist like me), Northwestern has financed the expansion of two multi-million dollar facilities through the sale of small slivers of the royalty stream (if memory serves, they sold 7% for ~$600 million in one transaction). So while it should probably be on the list, it’s hard to argue that it’s inclusion furthers the argument that the public shoulders the cost of these drugs. Lyrica is one of the best selling drugs of recent memory and will ultimately pay for virtually everything at Northwestern for years to come.

  30. Mike says:

    You could expand the thalidomide class of drugs. Thalomid, Revlimid and Pomalyst were all licensed from Boston Children’s hospital to Celgene.

Comments are closed.