Skip to Content

Drug Development

Scaling Up Arteminisin

A recent article in Science illustrates a number of points about drug development and scale-up. It’s about artemisinin, the antimalarial. Peter Seeberger, a German professor of chemistry (Max Planck-Potsdam), has worked out what looks like a good set of conditions for a key synthetic step (dihydroartemisinic acid to artemisinin), and would like to see these used on large scale to bring the cost of the drug down.
That sounds like a reasonably simple story, but it isn’t. Here are a few of the complications:

But Seeberger’s method has yet to prove its mettle. It needs to be scaled up, and he can’t say how much prices would come down if it worked. Using it in a large facility would require a massive investment, and so far, nobody has stepped up to the plate. What’s more, pharma giant Sanofi will open a brand-new facility later this year to make artemisinin therapies based on Amyris’s technology: yeast cells that produce a precursor of the drug. Although Seeberger says his discovery would complement that process, Sanofi says it’s too late now to adopt it.

The usual route has been to extract arteminisin from its source, Artemisia annua. That’s been quite a boom-and-bust cycle over the years, and the price has never really been steady (or particularly low, either). Amyris worked for some years to engineer yeast to produce artemisinic acid, which can then be extracted and converted into the final drug, and this is what’s now being scaled up with Sanofi-Aventis.
That process also uses a photochemical oxidation, but in batch mode. I’m a big fan of flow chemistry, and I’ve done some flow photochemistry myself, and I can agree that when it’s optimized, it can be a great improvement over such batch conditions. Seeberger’s method looks promising, but Sanofi isn’t ready to retool to use it when they have their current conditions worked out. Things seem to be at an impass:

But what will happen with Seeberger’s discovery is still unclear. Sanofi’s plant is about to open, and the company isn’t going to bet on an entirely new technique that has yet to prove that it can be scaled up. In an e-mail to Science, the company calls Seeberger’s solution “a clever approach,” but says that “so far the competitivity of this technique has not been demonstrated.”
The ideal solution would be if other companies adopt the combination of Amyris’s yeast cells and Seeberger’s method, [Michigan supply-chain expert] Yadav says; “then, the price for the drugs could go down significantly.” But a spokesperson for OneWorld Health, the nonprofit pharmaceutical company that has backed Sanofi’s project, says there are no plans to make the yeast cells available to any other party.

Seeberger himself is trying to make something happen:

On 19 April, Seeberger invited interested parties to a meeting in Berlin to explore the options. They included representatives of Artemisia growers and extractors, pharmaceutical companies GlaxoSmithKline and Boehringer Ingelheim, as well as the Clinton Foundation, UNITAID, and the German Agency for International Cooperation. (The Bill and Melinda Gates Foundation canceled at the last minute.) None of the funders wanted to discuss the meeting with Science. Seeberger says he was asked many critical questions—“But then the next day, my phone did not stop ringing.” He is now in discussions with several interested parties, he says.

As I say, I like his chemistry. But I can sympathize with the Sanofi people as well. Retooling a working production route is not something you undertake lightly, and the Seeberger chemistry will doubtless need some engineering along the way to reach its potential. The best solution seems to me to be basically what’s happening: Sanofi cranks out the drug using its current process, which should help a great deal with the supply in the short term. Meanwhile, Seeberger tries to get his process ready for the big time, with the help of an industrial partner. I wish him luck, and I hope things don’t stall out along the way. More on all this as it develops over the next few months.

23 comments on “Scaling Up Arteminisin”

  1. CMCguy says:

    It is not only the Chemistry that would have to be totally retooled but the nature of the process routes would mean extensive overall and redo of Regulatory documentation. That sometimes is even a bigger barrier than the actual plant operations and is a reason why often marketed drugs can be made by a poor process that does not get replaced unless the cost can justify introduction of a new route.

  2. anchor says:

    Derek: Correct me if I am wrong. Is it not true that we are already seeing artemisisnin resistant malarial pathogens in many part of the World? If true, my point was the flow synthesis solves little or no purpose, though as a chemist I think it is cool.

  3. milkshake says:

    I also think it is about regulatory hurles. In a flow system, especially in serial multistep arrangement the bugs are harder to track and impurity profiles and process stress margins more difficult to characterize. Scalability is not such a big deal in this case – even if the precess does not scale they could do in parallel a bunch of smaller flow reactors this could stil be advantage if they telescope the whole process in that way

  4. Paul says:

    (You appear to have some problem with the html; the italics have escaped.)
    BTW, does Toray in Japan still use their photochemical process for making caprolactam? That was at the time the largest scale use of photochemistry. I was reminded of it the other day when I read about a shortage of the monomer for nylon-12, which can be made by a similar process.

  5. Draviden says:

    First generation malaria vaccine is in Phase-III development now and the initial results are very encouraging.

  6. Electrochemist says:

    I found the following tidbit about artemisinin of interest in the Science article: “…the current price [is] between $0.80 and $1.20 per treatment course.”
    When we are talking about a western company like Sanofi making a semi-synthetic drug under cGMPs, I have to wonder how much more cheaply it could really be made? I wonder what percentage of the cost of the treatment is attributable to API costs?

  7. barry says:

    artemisinin-resistant malaria is–at least in part–caused by using sub-clinical doses of artemisinin. that in turn is often driven by the cost of the drug substance. Amyris’ goal was to make (synthetic) artemisinin so cheap that no one would be tempted to admnister the sub-clinical doses that cultivate resistant organisms. Of course it may not be too late.

  8. Myma says:

    If they are about to open the brand-new facility, the good professor is about a good 5 years (give or take) and a couple tens of millions of dollars too late for this round.

  9. Anonymous says:

    @6, plenty of cost sharing by various organizations brings down the final treatment cost.

  10. Anonymous says:

    It is my understanding the the price fluctuations of artemisinin are based on growing seasons; it is cheap after the Artemisia annua harvest and becomes more expensive until the next. The Sanofi collaboration represents a good solution to the pricing by introducing a difference source of drug to smooth over the supply bumps. Seeberger’s method could have excellent value, but has the disadvantage of showing up to the party a bit late for the first round of improvements.
    @2 It is true that resistance to this drug has been observed; It is also the case the artemisinin monotherapy is not a good way to permanently cure a patient. The WHO and other organizations recommend combination therapy with artemisinin and another antimalarial with three highlights (1) The artemisinin lowers the parasite load and suffering of the patient very quickly, (2) the second antimalarial drug kills the remaining parasites that artemisinin can’t seem to get, and (3) combination therapy raises the barrier to resistance. The true value of artemisinin is the speed of which it relieves patient suffering; a lot of patients do not present themselves to a clinic until they are hours from death.
    @5 A malaria vaccine is the ultimate goal for control of this disease; unfortunately, the data from all the clinical trials I have seen have a disappointing level of protection for a large price. After the challenge of efficacy and cost is addressed by vaccine technology, the next hurdle will be resistance. Again, I think we should be funding malaria vaccines heavily, but the realist in me believes that we will need small molecules to bridge the gap for at least a decade.

  11. metaphysician says:

    #10- Has anyone ever created a successful vaccine against a parasite before?

  12. Anonymous says:

    @11- No

  13. BigSky says:

    @11 Actually they have. It was shown in the late ’70’s that irradiated sporozoites injected IV traffic to the liver where they initiate a protective immune response against subsequent natural infection. (Almost identical results have been shown with genetically modified sporozoites instead of irradiated).
    The problem is that the ‘method’ is so burdensome that it will only be a niche tool for specialized applications like the military… if it launches at all.
    Seeburger is just behind the curve, unfortunately. A combination of sunk costs, low prices and institutional inertia will prevent adoption by Sanofi while chocking the air supply from any alternative approaches if they get theirs cranking.
    They also appear to be cutting off the air supply to the vaccine world too because, GSK RTS,S malaria vaccine fail aside, the net investment in non-academic malarial vaccine work is pretty close to zero.

  14. A Different Derek says:

    Re: comments 1 and 3 – artemisinin is not an API but only a raw material in the same way the material now extracted from plants is, so the regulatory burden should be minimal. The actual ACTs used contain an assortment of artemisinin derivatives and different classical antimalarials

  15. R. Friary says:

    Seeberger’s best course of action may be selling to Sanofi the rights to his process innovation.

  16. processchemist says:

    @ 14
    Artemisinin is the API and the continuous process starts from the same intermediate that the fermentative process is producing. As CMCguy and milkashake are saying, the issue here is regulatory. Impurity profile, process validation and so on – academia usually ignores these topics.

  17. A Nonny Mouse says:

    it should be noted that S-A is also developing Ferroquine for combination with the artemesinin. Ferroquine has no cross resistance seen with other chloroquine type anti-malarials. A unique drug at the moment with the incorporation of the ferrocene group which contributes to its profile.

  18. Jose says:

    Actually, the irradiated sporozoite vaccines had highly marginal activity which disappeared after several months at best. The uber-issue is we do not understand really anything about how natural immunity to malaria works (or doesn’t). There has never been a long term, validated vaccine against any multicellular organism, and certainly not one against one with as complex a life-cycle as P falciparum (let alone the other 3, or 4, human malarias). The most recent RTS,S results were sub-optimal, and the long-term followup? Well, we’ll see… (and then there is the issue of 4-5 billion syringes for three rounds).

  19. Vader says:

    Artemisinin, eh?
    So all those Dutch artists getting stoned on absinthe may have been on to something?

  20. A Different Derek says:

    @ 16
    Beg to differ. Artemisinin as such is not used in any of the WHO approved ACTs, which mostly contain artesunate and I think there might be a dihydroartemisinin one out there too. Sanofi’s position, which has already passed regulatory muster, is that artemisinin is a raw material, not an API. Re where the continuous process starts, it is dihydroartemisinic acid, NOT the artemisinic acid produced by the fermentation route, so to start the continuous process one would need: a) access to fermentation product and a chiral hydrogenation to the correct dihyroartemisinic acid isomer for the subsequent cyclization to artemisinin; b) access to a good source of plant extracted artemisinic acid (like artemisinin itself, it is found in some accessions of Artemisia annua plants, sometimes in higher amounts that artemisinin), but you are still dealing with extracting from a plant source AND you still have to hydrogenate or c)find a natural source of the correct dihydrodroartemisinic acid – also a natural product found in varying amounts in different Artemisia plants – saves the hydrogenation step but doesn’t solve for the whole extraction from plants thing

  21. Anonymous says:

    To my knowledge Artemisinin can be used alone, but in return leads high rate of recurrence and if that is so ,other drugs are required prevent return. Worldwide WHO has recommended artemisinin combination therapies (ACT) to be the first-line therapy for P. falciparum malaria .Combinations therapies are effective because the artemisinin component are known kill the majority of parasites at the beginning of the treatment, while the more gradually eliminated associate drug clears the other remaining parasites. The goal of making a semi synthetic artemisinin so cheap I think is a good idea that in turn will avoid resistant organisms. I also think that the First generation malaria vaccine which is in the Phase-III development should be looked into.

  22. Anonymous says:

    Contrary to what some of you have been told Artemsia annua is a extremely robust and easy to grow plant, that can be grown in a very wide range of conditions and still yield well. Now with new cultivars with higher levels of goodies easy to extract, one surely needs to scratch ones head and wonder what went wrong? in the last 30 years or so we have been aware of this drug?
    Was it because chemheads with not the slightest knowledge of botany were in charge?, was it WHO being misled?
    Whatever it was that stopped the upscaling of the use of this wonderful botanical laborartory machine “Artemesia annua” and forced us down the long and twisted path of synthesis being the only way to scale up, this blantant lie has resulted in the loss of potentially milliions of lives
    Here we have green chemistry, and in the age of searching for new more energy efficient (calories in to drug out) we have it in our hand and ignore it head buried in work thinking like a man wanting to reinvent the wheel.
    Depressed rural sectors are looking for new markets and crops, and to reduce imports into struggling developing countries and with a fraction of the cost you have spent so far you could of had an extant and thriving Artemsesinin Cottage pharmaceutical industry in just about every country that needed it. So that they were not depednent on our kind benevolent souls with our on off game in the labs,
    I am glad the stalmate was broken to a degree by the French whom had the decency to acknowledge the Chinese and try to work togwether with them on plant sourced Artemisinin, and that this has encourage some others, despite some grumblings at WHO.
    I feel many of the people, including Derek, writing here are good people that on hindsight if they knew the full facts in this case and had a understanding of the horticulture and the lies used to push things down the synth path as the only way, full stop with WHO, they would be appalled at the outcome in increased human tragedy.
    I am not saying go only the plant route either, we all know a system designed for resilience needs back up pathways
    Surely the use of Occums Razor would of led to the plant path being set in place as the first phase? to save the lifes of countless people, and the means to this drugs production put in the hands of all nations that needed it?
    Then if anything better came along, if and when it comes along (which in my opinion it has not yet) then great! we have backup or perhaps a new step forwards.
    Could it have been prevented if some knowledgable horticulturists and not only chemheads were part of the decision process?, I think there is a need to change a mindset, we need to stop looking at plants fungi (and not just microbes) as crude unreliable unpredictable complex primitive oddities. In my opinion we will need to look more and more at them as highly precise and accurate (which they are given the tolerances of homeostasis survival immunity and growth). They can be powerful highly efficient solar powered machines driving complex laboratories, that with little work or input by us produce some products or highly resolved precursors.
    We may have to go down this path sooner rather than later if peak oil has passed.
    Okay i may be wrong, I am a horticulturist and carry that bias no doubt? but I welcome any constructive challenge to what I have written and thanks Derek for providing this blog and chance to air a view that may be a bit contrary
    Some of which over the last 20 years appeared to be sitting on their hands on this
    Not some big centralised concern
    this polant is in essence a solar powered Artemesinin producing machine. No modern lab operates anywhere near as effeciently or effectively and accurately as these botanical machines.
    Perhaps because you guys are all seemingly to me coming at this from the chemhead route and not the botanical head route.
    I think some of you guys need to take a step back, there has been endless lies and perhaps a bit of cloak and dagger work (ie blantant theft of Chinese intellectual property, and you wonder why they dont care about bootleg copies of windows on the streets)

Comments are closed.