Why Synthetic Artemisinin Is Still a Bad Idea - A Response to Rob Carlson

Background: At a recent Synthetic Biology Conference in Cambridge UK, Synthetic Biologist Jay Keasling announced that the consortium he was working with now intend to replace the entire global supply of artemisinin (an anti-malarial compound) with their new synthetic-biology derived version. I was at the conference and challenged some of Jay Keasling's claims first in person and then in The Guardian. In response prominent Synthetic Biologist Rob Carlson leapt to Keaslings defence arguing that Keasling's synthetic artemisinin was a better bet for tackling Malaria than existing botanical supplies grown by farmers. He asked ETC to reconsider our opposition based on his blog article. We investigated his arguments and below is our response to Rob.

You can read Rob Carlson's original blog article (and this response) here: http://www.synthesis.cc/2013/05/the-economics-of-artemsinin-and-malaria....

You can read my original Guardian article here: http://www.guardian.co.uk/global-development/poverty-matters/2013/apr/12/synthetic-malaria-compound-artemisia-farmers




Hi Rob,

So you specifically invited me "to consider the facts and think through the problem" regarding artemisinin. I have done so and consider that your continued excitement about synthetic artemisinin is still misplaced. I'll address that below but first 2 quick observations about the first half of your post:

1) A historical note: Have you thought about how some of the problems you think might be addressed, theoretically, through synthetic biology emerged from the large scale and rushed application of a technology that wasn't properly understood at the time? Large scale burning of coal – and then other fossil fuels – for motive power seemed an industrial good for much of the past two centuries until scientific consensus belatedly supported what Svante Arrhenius first suggested in 1896: that the resulting CO2 would warm the atmosphere. Only more recently has the additional problem of ocean acidification become apparent. The cause of honeybee decline might still not be fully understood, but as you say, the recent EU moves against neonicotinoids are sensible and a part of a belated recognition that the “new” synthetic chemistry, now 150 years old, has massively harmed biodiversity (a story long resisted by the synthetic chemistry industry -- even today). What this speaks to is the need for real precaution and humility in the roll out of new technologies and not jumping to speculative claims about what a technology might do without equal enthusiasm for finding out the less desirable effects new technologies may have. Using synthetic biology to correct the mistakes of previous technology introductions might be a 'swallow the spider to catch the fly' kind of response (and you know what happened to that old woman…)

2) Your argument that we need to develop conservation-relevant applications of synthetic biology in order to understand if synthetic biology techniques might prove to be a plan B for dealing with biodiversity loss is strikingly similar to what the geoengineers argue. They want to develop the technological hardware to alter the Earth’s climate to understand if we might need it. It’s also not a million miles from what the biodefense guys argue when they say that we ('the good guys') need to develop bioweapons to understand them in case they ever get used against us by the 'bad guys'…in the meantime the technologies get developed, owned, and acquire their own momentum and lock in, regardless of the consequences.

As for artemisinin:   

So as I understand it, you lay out two arguments for the value of synthetic biology-derived artemisinin: 1) it will make malaria drugs cheaper and therefore that’s best for combatting malaria overall 2) Artemisinin farmers are caught in a cartel that works against their interests –  as people living in at-risk malarial areas – by keeping Artemisia-based drug prices high. Both arguments basically accept that the synbio-derived version will be cheaper.

In your post the revealing comment is that you say this second argument “you only realized while  writing [your] post,” which leads me to suspect it’s not based on much actual investigation into the politics or economics of the market for Artemesia, but is  a bit of intellectual theorizing on your part, Ie, it sounds like a good argument in the abstract, but how does it match the reality?

For background, it’s worth looking at the various arguments that have been put forward over the years for why Syn Bio-derived artemisinin might possibly be a good thing to develop.

First there is the fundamental purpose for the project, namely, a way to develop an industrial platform for producing isoprenoid chemicals. That’s the honest reason we now have semi-synthetic artemisinin in the market. – Vince Martin, who did the original work, is fairly candid in saying that he and his colleagues were simply looking for a candidate molecule that could attract funding to their isoprenoid platform work and that artemisinin fit the bill perfectly – synthetic artemisinin had a compelling story, a funder with deep pockets and a great PR angle. They went to the Gates Foundation and leveraged philanthropic money to found a startup company (Amyris) and, once the microbial platform was built for artemisinic acid, Amyris turned its attention back to what it was really interested in: other isoprenoids such as biofuels, high value compounds such as rubber, patchouli, squalane etc...

The ‘story’ of SSA (semi-synthetic artemisinin) has changed over the years:  Jay Keasling first claimed that it would overcome environmentally damaging means of producing artemisinin, but dropped that argument quite quickly. He then claimed it would fill in the shortfall in artemisinin production, but dropped that argument when it became apparent there was overproduction of botanically-derived artemisinin. He and Amyris then claimed synthetic artemisinin could be used to smooth out the boom and bust cycle –  which was a bit more nuanced an argument, but not ultimately convincing since it imagined a pharmaceutical company refraining from 'firing up its bioreactors' in years when botanical production was assured, even though the single source would be  advantageous (ie, not having to source from different botanical producers). It also ignored the fact that there was already a concerted effort to overcome the swings in production through the A2S2 (Assured Artemisinin Supply System initiative) and that those swings have been largely overcome. Finally, just recently Jay Keasling has admitted that, yes, the aim (his aim anyway) is to replace all of the botanical production but now the rationale is that this is a means of undercutting the economics of production of monotherapies that may be responsible for emerging artemisinin resistance. (That’s unproven, by the way). That’s the way a drug company argues, ie, if you hand us control of the entire market we can stamp out less savory practices and the price you pay for that is that we, the drug company, end up with a monopoly.

A few observations on where I disagree with your analysis:

1. There is no 'pricing cartel.' I wonder if you relied on Keasling’s language [from Cambridge?], which, as far as I can tell, refers to the A2S2 (Assured Artemisinin Supply System) initiative. A2S2 is a very sensible process in which all the players in the production of ACTs (funders, growers, processors, drug companies) have been working together to ensure a stable supply of botanical artemisinin to deal with the swings in prices that existed before there was a real structured market. It’s a sensible way to ensure steady supply of a life-saving drug, which is as good a reason as any for all the players in the market to collaborate. It has been a very successful endeavor and botanical production is now meeting demand (and indeed demand is expected to drop in the next little while for technical reasons to do with monitoring). From my conversations with players in the supply chain, Jay’s 'cartel' and 'price fixing' characterization has raised a lot of hackles and is inaccurate, though those involved suggest he may not understand the A2S2 process. Or do you have other sources for your ‘cartel’ claim? As cartels go, they are not very successful so far at pushing up prices (see below)… they could use some coaching from OPEC, or Monsanto.

2. There is no evidence that synthetic artemisinin is or will be any cheaper than botanical artemisinin. Its real advantage to drugmakers is not so much price as the fact that it can be more easily controlled when it is brewed from a single source, so the drug companies don't have to maintain complicated sourcing arrangements. Initially the SSA consortium promised $100 per kilo artemisinin, which would indeed have been a killing blow to natural artemisinin but they didn't actually realize that promise. What they realized instead was around $380-$420 per kilo and that is after their production and development costs being heavily subsidized by Gates, etc. We don't yet know the actual cost of SSA. It will be determined over time by the cost of sugar, the cost of energy and the internal costs of Sanofi Aventis and others.

The great misconception about natural artemisinin pricing is that it has been very high and that is higher than what SSA is offering to the market. The average price is actually quite low. If you take the 745 metric tons of artemisinin that has been imported into India over the last 10 years for processing by the large Indian generic pharmaceutical companies, the average price of the last 10 years is $370 per kg…a price point which is, in fact, lower than what the SSA promoters are targeting. Nor is the fair price of botanically-derived artemisinin necessarily going to always be above SSA in the future. New breeding work on artemisia is developing varieties in which the leaves have higher yield of artemisinin per leaf. (Indeed Gates are also funding some of that work.) That combined with improvements in extraction efficiency and denser plantings could mean increased yields so that  producers may yet be able to compete on price while still realizing a livelihood -- maybe even do better in keeping price down -- who knows? That said, even if botanical artemisinin ends up cheaper Sanofi may yet choose to keep their own in-house SSA source because it is simpler. Note that artemisinin price is only 40% of a final ACT drug price. So final drug prices are at best only partly determined by ingredient costs.

3. The idea that taking 100% control over global artemisinin production and driving down price through synthetic production will drive out the monotherapies market is a convenient argument for a pharmaceutical company looking to justify monopoly behavior but is unrealistic as an anti-monotherapy strategy. In part, this is because there is no proof that SSA will be cheaper (see above). A2S2, for example, already has mechanisms to delist anyone who is discovered to be selling into that other market. Realistically the 'other' market outside ACTs production (including monotherapies) will always be there and will always find ways to get hold of cheap artemisia leaf, which, from some quarters, is quite elastic in price. Consider what happens to the growers, extractors etc who formerly sold to Sanofi but have just discovered Sanofi no longer wants their product – where do you think they are likely to sell to now that Sanofi is no longer buying their product? Far from starving the monotherapy producers, this move may be a boon for that 'other market'.

4. You talk about 'a few thousand farmers'. The most knowledgeable people in the artemisinin supply chain peg the number of farmers at about 100,000 and point out that the social impact of lost livelihoods is about 3-5 times that (account for families).

5. The graph showing that farmers in China and Vietnam could switch to potatoes was used by Jay in his talk in Cambridge in the absence of context. It was originally used as a part of a report by the Boston Consulting Group showing how artemisia farmers need ways to improve and augment  their incomes, not to look at feasible alternatives to artemesia. Many artemesia farmers in China do grow potatoes already (it turns out it’s an excellent break crop for artemisia) but then again many regions where artemisia grows cannot grow potatoes – that’s true in parts of China and very true in East Africa. Also artemisia farmers already do grow food eg, Madagascan artemisia farmers rotate their production with rice since they have two growing seasons per year. It’s not that artemisia farmers aren't growing food – they are. Artemesia is a cash crop ON TOP of food production – it’s the crop that keeps them out of poverty. Take away artemisia and they will still have some food but little to no cash.

What is perhaps most worrying is that Jay's pronouncements about intending to take over the entire supply of artemisinin and claiming to undercut botanical production may already be driving farmer planting decisions right now as well as investment decisions. As Malcolm Cutler of A2S2 warns in the Nature article that you link to: “If it’s brought in too fast it could create huge shortages, because people will stop producing the natural stuff.” If farmers decide not to plant because they fear they can't compete with SSA, we may be in for shortages in 2015 – that really would be a tragedy. If there isn't enough artemisinin for ACTs people may die. Introducing synthetic artemisinin the wrong way will  not only hurt the farmers who lose their livelihood, but could also hurt malaria sufferers.. Indeed I’m told there seem to be some lessons from previous misguided communication led by CHAI (Clinton Health Access initiative) a few years ago when their unrealistic attempt to lower ACT prices led to a complete stoppage of Artemisia plantations in China, and in fact, contributed to the more recent shortage and price surge.

From my conversations with those in the supply chain it seems there is now an urgent need to clarify (and most probably rectify) the public communication around SSA introduction. Are Jay Keasling’s public statements his own or does this reflect the position of the SSA consortium? Has the consortium really decided on an all-out take-over of the market for reasons that are removed from the real needs of malaria sufferers and farmers?