Competition Compounded

ArsTechnica: Turing Pharmaceuticals, the company that last month raised the price of the decades-old drug Daraprim from $13.50 a pill to $750, now has a competitor.

Imprimis Pharmaceuticals, Inc., a specialty pharmaceutical company based in San Diego, announced today that it has made an alternative to Daraprim that costs about a buck a pill—or $99 for a 100-pill supply.

Good news. Competition is working. But I was puzzled. In Generic Drug Regulation and Pharmaceutical Price-Jacking, I argued that competition was slow because the FDA makes it expensive and time consuming to get a generic drug approved for sale in the United States. Was I wrong about the difficulty of generic entry? No.

The drug that Imprimis Pharmaceuticals is selling is not FDA-approved. A bit of background is in order. Even today some drugs need to be tailor-made. A patient, for example, might not be able to swallow a pill so a licensed pharmacist may create for a specific, individual patient a small batch of the drug in liquid form. A pharmacy that does this kind of work is called a compounding pharmacy.

Compounding pharmacies have a long and tendentious history with the FDA. The FDA has always claimed that a new drug is a new drug, even one created only for a specific individual. Thus, the FDA has always said that it has the right to regulate compounding pharmacies just like manufacturers of new drugs. In practice, however, the FDA allowed the industry to proceed with little regulation.

In the 1990s some compounding pharmacies began to create large batches, especially of drugs in short supply, as a way of avoiding the FDA process. The FDA worried about quality control, however, and it re-evaluated its traditional hands off approach. A political battle then ensued in which Congress and the Supreme Court also had their say. In 2012, fungal meningitis outbreaks caused by poor quality control at the New England Compounding Center brought these issue to public attention and resulted in greater regulation of compounding pharmacies, albeit with clearer regulations on when a compounding pharmacy may sell large quantities.

Imprimis Pharmaceuticals did not apply for approval to sell a generic version of Daraprim. As I argued earlier, that would take years and cost millions of dollars. Instead, it is doing an old-style end-run of the FDA process by offering its alternative under the compounding pharmacy laws. That means that it can only sell to order, on a patient by patient, prescription by prescription basis. Since Daraprim is not widely used this may work. Indeed, I hope this end run works but my reading of the act is that compounders can only supply drugs in large quantities if they are on the FDA’s shortage list. Perhaps the FDA will look the other way, however, in order to send Turing and similar firms a message.


It is supremely foolish to think FDA should just allow any generic drug on the market as soon as it is made, with no review to ensure it is safe and will be manufactured under sufficient quality controls, which Alex is always advocating for.

But I will credit him for debunking some of the news coverage making it seem that this new drug is now widely available as a generic to replace Daraprim.

Given that your way of doing things is resulting in spiraling costs, reductions in supply and a market place dominated by predatory monopolists, have you considered that you could possibly be wrong?


What evidence would be required to get someone to reconsider their position?

Are you already convinced that drugs should not be reviewed to ensure safety?What level of trade offs would you accept? How many dead?

How many dead are you willing to accept from lack of drugs on the market, or in this case, lower compliance from lack of drug competition?

Do you know how many people current policy has killed? If you don't have a number because you can't even evaluate the current landscape, you cannot convince someone there should be a wholesale change.

Also, please point to the generic manufacturer who has publicly stated they are ready to mass market Daraprim for a lower price.

Seen versus unseen, displayed in person.

What do you mean by my way of doing things? Having FDA review drugs for safety? It is absurd to suggest we not do that because of one drug that, what, 4000 people take and has becaome too expensive.

There are a few obvious changes we could make the system. 1) Set rationale drug pricing policies like the rest of the world does, rather than subsidizing every other rich country. 2) Ban 5000% increases in drug prices?

Tell me what you would change to fix these problems

Mike brings up the bigger problem. You need to answer that one.

No he doesn't. The onus is on him to tell everyone how many people we are killing under the current system. Surely, it is easier to evaluate the current landscape it is so terrible and needs to be done away with.

What are your thoughts?

How can you possibly tell how many people we are killing? Probably tens of thousands at least. How is it "surely easier" to evaluate how many people are killed by medicines NOT approved than people killed by medicines that ARE approved?

It's easier, because it is actually possible to count dead people in the current reality, while I would argue it is pretty much impossible to estimate the number of dead people in some counterfactual reality where the FDA safety standard is no more.

Given how many in Alex's tribe are hankering for elimination of the FDA, I am sure they can provide some number of deaths attributable to the current FDA review standards.

Lack of review by the FDA does not mean that the generic drug will be harmful or ineffective. It only means that a burecraut has not looked over their shoulder with a clipboard prior to them going into production. Producers still have every incentive to be safe. They could be ruined or go to jail if they are not.

The FDA does not do a prior review of every new food manufcturer and you still feel justifiedly safe.

Finally, the requirement of FDA prior approval actually makes drugs less safe, as manufacturing has to be done on a "per order" basis, without the incentives for investing in better facilities and procedures that may scale production up and allow for even safer systems.

I don't think you understand how generic drug review works. The main focus is on how the product will be sourced and manufactured, whether they have in place processes to ensure quality, etc. They do not require any human or animal testing. The molecular entity itself has already been studied in humans and used for a long time--it is proven safe.

Do you know how many warning letters FDA issues each year due to manufacturing quality violations they find?

Which drug makers have gone to jail recently for making unsafe products?

There are no such thing as "new foods". I do think they should review new chemical entities that they place in food as supplements.

Your last point is malarky.

Don't they require that you show the generic is as effective as the original?

They don't have to do clinical trial, or evaluate the effectiveness or clinical outcomes of the generic. They just have ensure generic's active ingredient is available at the same level as the original product.

Jan - are you sure that generics don't have to be tested in humans?

In Canada, generics producers do clinical testing all the time. Does that not apply to the USA?

Automatically accept Canadian and European approvals for the US market for say, a five-year period while waiting for domestic approval mechanisms to be fulfilled, with options for renewal of the five-year period. And allow tariff-free, quota-free access to the US market for these generics.

Oh okay, so shut down all medical innovation and limit price increases to 4999%, sounds like you have thought t his through pretty well.

Yeah, Cliff. Shut down all medical innovation. Or how about pay for drugs based on their value, rather than what the manufacturer charges. It is completely idiotic we don't do that.

Cliff, the cases of Daraprim and drugs like it are clearly special. There is no reason to think that measures to ensure competition for 75 year old (or however old Daraprim is) drugs with expired patents will do much to stifle innovation.

It's not just one drug, there are hundreds of drugs that suffer from shortages or greatly increased cost due to FDA policies. If you are a generic drug manufacturer and asked the FDA for a license to manufacture Daraprim you would be denied because the FDA will only allow one manufacturer for certain drugs. It has nothing to do with safety of efficacy and it is only a problem because the FDA grants monopoly status to drug companies. That's what Martin Shkreli paid $55 million for, a government sanctioned monopoly, the right to sell Daraprim in the US.

That's why they should just allow imports. It's not like there's any economic contribution to one group of Americans shelling over loads of dough to some other Americans for doing nothing useful, when it can be bought for a pittance elsewhere.

That's completely untrue, they never allowed just one manufacturer for generic.

Perhaps it's you who ought to invest in a little epistemic humility. You seem to be committing the classic error of attributing enormous power to your ideological opponents and pinning all of society's woes onto them.

Requiring generics to be tested and regulating quality control is SOP in every developed country in the world, typically with success. Of course you're welcome to argue against it, but it would be silly to treat it as some crazy unworkable scheme that Jan and his liberal buddies came up with.

Looks like I'm too late. Jan's comment is a better reply than mine. Ignore mine.

Isn't the problem a bit more complicated than that? If an increase in regulations leads to drug shortages, then it's not obvious that welfare is improved under this system of regulations. No one hears about deaths from drug shortages but everyone will certainly hear about deaths from a contaminated facility. That being said, I agree that the standard libertarian refrain about competition ensuring drug quality does not necessarily work here, because the FDA holds a monopoly on quality control. So without a private and transparent system of drug quality assessments, you risk having NECC-type incidents.

For sure. Testing does take time and cost some millions of dollars. Normally those costs are small enough relative to the amount the drug will bring in that they're worth pursuing nonetheless, so most of the time this is a policy that works well.

Obviously that's not happening in this case because we're dealing with such a small drug. So what do we do about it? We could a) completely dismantle the testing and quality control apparatus that we have (the libertarian solution), or b) we could simply regulate the price of these particular drugs (the more leftist solution).

My opinion is that (a) is throwing the baby out with the bathwater, using a sledgehammer to kill a fly, etc., while (b) is probably better. Yet the above commenter seems to think Jan is the one who needs to change his whole outlook and adopt (a) as the obvious choice, despite the fact that (a) is the position that's completely off the reservation as far as the Overton Window is concerned.

Why does it need to be approved 20 times over, once for each market?

@Jan would you support for example, the FDA allowing the sale of any drug reviewed and allowed for use in Germany, The UK, France, The Netherlands, Sweden or Norway with their own review?

In some cases, like with Daraprim and in cases of true shortages, I would.

Should the same criteria that are used for drugs be used for surgery? Would this end in a ban of back surgery?

Apple and oranges, but it might be a good thing, because the majority of back surgeries are not effective.

"It is supremely foolish to think FDA should just allow any generic drug on the market as soon as it is made"

Why, since the drug the generic is based off of has already been through extensive FDA testing?

Bioequivalence. In first year organic chemistry, you learn about so-called "left" and "right" versions of molecules (think of how lettering looks the wrong way when looking in the rear-view mirror for an intuitive idea of what this means), and even if it passes a chemical test for equivalency, it might have more "right" and "left" than another production method.

I believe he does not mean that it should be re-tested, but the facilities and production do need to be.

Isn't that still so expensive that no one wants to make the drug?

Or am I misunderstanding Jan?

Yes. They don't need to retest it in humans. Lab analysis is good enough for ANDAs.

"Perhaps the FDA will look the other way, however, in order to send Turing and similar firms a message."

The FDA has as their mission to distort, if not abolish, the market. Someone comes around that introduces some semblance of a market. Now the FDA are expected to look the other way?

Yes, one can hope. But don't hold your breath.

"The FDA has as their mission to distort, if not abolish, the market." Really??? Any evidence that what you say is not gibberish?

Ahhh, an oppressive government political bureaucracy with "rights".
How quaint.

Where do these alleged FDA rights originate. Certainly not from the U.S. Constitution.

Federal politicians and bureaucrats have absolutely zero legal or ethical right to dictate what Americans may or may not choose
ingest/inhale/inject. It's all a highly arbitrary and harmful power game.

FDA's entire premise is to forcibly intervene in the voluntary choices of buyers & sellers (the market), substituting the arbitrary choices of unaccountable government bureaucrats for those of supposedly free Americans.

is it troll day today or what???

No rights are absolute. The courts are expected to exercise good sense in addition to good reading comprehension.

What I still don't understand is where Imprimis is actually getting the drug from. Do they synthesise the chemical? Do they buy it from India or from a veterinary supplier or something? Can anyone enlighten me?

Apart from that, this sounds like a decent compromise actually. Of course we should have much stricter regulation on drugs to be sold to millions of people than on drugs to be sold to hundreds.

This is all done via compounding pharmacies, so they get in the components (pyrimethamine and leucovorin), which I do not believe are in short supply, and produce the equivalent of Daraprim, which is the drug for which Turing jacked up the price.

It's simultaneously an end-run around Turing and an end-run around FDA's policies on generics.

And, also, continuing proof that competition works wonders when not impeded too much by government.

Turing, in essence, is not so much a pharm company as a company that takes advantage of how the FDA restricts competition. And Imprimis, in turn, found a way to compete anyway.

Thanks for the reply.

I'm still a bit confused though. Are you saying that there are companies already manufacturing Pyrimethamine wholesale? It doesn't seem to have other uses though. Maybe for export? Did Turing even make it themselves, or do they just own the name and the paperwork, and maybe a packaging plant?

Or is the compounding pharmacy Imprimis in the business of chemical synthesis? I thought they just mixed powders or dissolved things in alcohol, but perhaps I am wrong. From here wiki/Pyrimethamine#Synthesis it doesn't sound all that hard, although I can't really tell.

Yes, it's made all over the world or the chemical could probably be synthesized by a chemistry student. The tricky part isn't getting your hands on pyrimethamine, it's having approval to legally sell it in the US for human consumption.

This compounding pharmacy found a way around the legal restrictions.

You ask an excellent question, which I myself asked (tweeted) of Imprimis. Their response tweet: "Our source for the active pharmaceutical ingredients (APIs), including pyrimrthamine, come from FDA-registered suppliers of quality pharmacy compounding products for North American [sic]."

There are numerous companies around the world that supply APIs for drug manufacturers (these manufacturers have to get the raw, bulk chemical material from somewhere, if they don't produce it themselves). As an example, this link provides a 2010 WHO survey of international sources for APIs: Many of these sources are overseas and, as you can see, some have been certified by the US FDA as being "GMP compliant." Also many of these (at a glance) are in India. I suspect, but I certainly do not know, that Imprimis gets its bulk-supplied pyrimethamine from one of these FDA-certified sources that's legally allowed to export its bulk API product to the United States.

BUT for all we know, Turing does the same thing. As far as I can tell, Turing's Daraprim, aka branded pyrimethamine, is actually manufactured and distributed by Amedra Pharmaceuticals (which was acquired by Impax, from which Turing bought the marketing rights to Daraprim). Notice on the PI for Daraprim (at this link through Daraprim Direct,, that "Amedra" is on the document, not "Turing." So it appears that, in Turing's acquisition of the marketing rights to Daraprim, it left the manufacturing of Daraprim with Amedra...Because, why change a manufacturing process for your product that's already in place? And, for all we know (once again), Amedra gets its bulk API pyrimethamine to make Daraprim from the same or a similar source as Imprimis. Namely, an overseas API supplier that's been FDA-certified as GMP compliant.

Until further information on the supply chains for API pyrimethamine is forthcoming, my hunch is that all roads lead to India. And I suspect for many of the APIs that we consume in FDA-approved prescription drugs (especially generic drugs), this is the same situation.

So if the bulk API supplier of pyrimethamine is similar (or the same!) for Turing and Imprimis, the job of Imprimis is to not f*ck up the compounding of their versions of the drug, by ensuring sterile and safe practices in compounding at their US laboratories/pharmacies.

More follow-up for "improbable" (if anyone's still reading this comment thread).

Turing and Imprimis may not have to get their APIs from India, after all--unless there's a some really, really compelling cost difference, which there may be. For instance, Spectrum Chemical Manufacturing Corp offers 1 kg of USP pyrimethamine for $617.25, or for about 6 cents per 100 mg (if I'm doing the math correctly). The per-mg cost is reduced by more than 50% if you buy 25 kg (<3 cents per 100 mg). Turing/Amedra's Daraprim tablet is 25 mg, so the pyrimethamine that goes into it would cost them about 0.7 cents per pill (if they bought it from Spectrum--and I have no idea if they do). According to Spectrum's website, they have "FDA registered drug establishments" in California and New Jersey that "operate under current Good Manufacturing Practices (cGMP)." So anyway, that's a scenario in which Turing/Amedra and Imprimis could get their USP pyrimethamine from the same US source and pay less than a penny per mg for it.

I thank bmartinmd for the extensive additional details. I was wondering about some of those details myself.

Another follow-up to my follow-up. In re-examining the prescribing information (PI) for Daraprim, it identifies Amedra as the distributor, but there is no manufacturer listed--which I find weird. Maybe there's an innocent (?) explanation for the lack of explicit information on Daraprim's manufacturer, but in the short interim since I originally posted here, I haven't been able to determine who actually manufactures Turing's Daraprim for US distribution/marketing. Amedra, which is a subsidiary of Impax, appears to be a drug distributor solely--not a manufacturer. Turing certainly isn't the manufacturer; they merely (merely!) have the US marketing rights to the drug. Yet, the Daraprim for US distribution must be made somewhere.

In looking at earlier Daraprim labels (where I can find them online), it shows that an outfit by the name of DMS Pharmaceuticals in Greenville, NC, made Daraprim for GSK (originally) and later for distributor/"labeler" Amedra (which was/is a part of Impax). I'm wondering if that supply chain is still intact, more or less: Namely that DSM still makes Daraprim, but now for Turing. But why wouldn't that information be listed on the PI?

It may be notable that DSM merged with Patheon last year. Perhaps that event has something to do with a change in the name of Daraprim's manufacturer or the possibility that the name of Daraprim's manufacturer is in a state of flux or the manufacturer of Daraprim itself is in a state of flux. I'm at sea here.

But what really should be a simple question to answer--Who makes Turing's Daraprim?--is becoming really hard to answer.

Are people really taking this to be a big win for the "markets" side of things? Has anyone, anywhere, EVER argued that drug prices were high due to anything other than monopolies granted via IP protection and safety regs?

The point is to argue that the safety regs are too stringent and the IP protection too restrictive. This is just a case of ignoring both completely.

I interpret Tabarrok's obsession with the FDA as evidence of his dedication to the cause, the cause of less government and more markets. After all, if we'd be better off without the FDA, then we'd be better off without government. Extremism in the defense of liberty is no vice. Or something.

Or maybe his dedication to providing more medicines to people who need them, for which he believes, and provides evidence and arguments, that more government and markets would be an effective solution. You might not think is evidence and arguments are convincing, but there's no need to attribute mindless fanaticism to what seems a pretty reasonable position.

"more government and markets" should be "less government more markets"

Of course, it could be that miracle cures would be available at low cost if the FDA would just get out of the way. Or it could be that that snake oil would be available at low cost if hucksters and medicine men didn't have to contend with the FDA. Both alternatives depend on an irrational belief in potions and elixirs. If Tabarrok's point is that markets will produce cheaper and effective pharmaceuticals because of the rational behavior of producers and consumers, he's picked a very odd subject. Desperate people are anything but rational.

I agree there is something like that going on. He takes very extreme, ideologically driven positions (see: open borders) because he thinks someone has to be out there pushing those views. But it's an academic exercise for him, and his positions are so extreme that he knows he will never have to suffer the consequences of them being implemented.

... and, of course, your own positions here are never ever ideologically driven nor extreme.

your assertions here are in no way academic or rhetorical exercises -- and you know you will directly suffer the consequences of your positions stated here

My opinions on those issues are very close to the status quo. That is by definition not extreme. Alex wants to open all borders and do away with the agency that ensures the safety of goods consumers spend 25 cents of every dollar on. That is objectively extreme, even if you think it is the best policy approach.

you neglected to address ideology in your positions.
are your views unaffected by ideology?
how do you objectively know/measure that these positions are "status quo" ?
how is it objectively impossible for the "status quo" to be extreme?

Good post. I never knew about these kind of pharmacies.

Also of note is the comeback of 23andme. Good news but my fear a chastened 23andme will not pursue the really interesting genomic information that they discover. There is a chill in the market.

For drugs which are supremely expensive in the USA and cheap elsewhere, I believe you are allowed bring back some number of months of supply from a foreign country, on your person.

I looked into the regs for this for Canada, while I was in India, and you are (or were at the time) allowed to bring back a 3 month supply for any immediate family member.

So for something at $1000 a pop, the cost of your plane ticket is easily covered, and why not take a short holiday while you're at it?

What a tragedy. What does approval by the FDA mean for a generic version? I'd guess "ask the manufacturer for their data/plans and we will sit on it for years and then OK it if we think you did a nice looking file."

The thing is, there is no pending generic for Daraprim at the FDA right now. It takes money to set up a manufacturing facility to make this drug and any manufacturer knows Turing will just lower the price of another generic of it does reach market. So there's not much invective to do this.

The drug is also available from Canadian pharmacies for about the same price, or you can buy it by the kilogram from Ali Baba.

I don't know him. Is his stuff good?

Alibaba = website that has a directory of Chinese manufacturers. The Yellow Pages, you might say.

Go ahead and order from someone you find there. Hope you got your own chemical lab to do composition and purity testing!

Imprimis Pharmaceuticals did not apply for approval to sell a generic version of Daraprim. As I argued earlier, that would take years and cost millions of dollars.

Hmmm, so after some digging around I saw that in the US about 2,000 patients use daraprim at any given time and the illness it is used for, toxoplasmosis, requires about a year of treatment.

So 2000 people * 1 pill per day * $1 per pill = $730,000. But before all this huffing, the price was $13.50 a pill. That's $9.8M per year. Not a lot in terms of the biggest selling pharmaceuticals but actually that's enough of a cash flow to spend a few million to verify your plant can actually produce the pill properly. And that is if we buy the claim here that it would take millions of dollars to produce the pill due to compliance with FDA regs. I'm not a chemist but from the wikipedia article it appears this is not a very complicated molecule. I suspect the manufacturing process would not be very complicated nor would the process of demonstrating you were achieving and maintaining quality control. It probably is not something you can do in your basement or garage, even if you were Walter White.

This doesn't seem as much a problem of health policy or even pharma policy as much as the problem of a natural monopoly. Manufacturing the pill has fixed costs. Those fixed costs are not covered if you have a lot of players in the market driving down prices. Hence the market veers towards fewer suppliers otherwise the price will fall below costs. Making it slightly easier for firms to enter the market may help things slightly but I suspect not that much. $730K per year doesn't give you a lot of room to cover fixed costs. The best thing that could happen here is if someone could find some other use for the drug to expand it beyond just 2K patients per year.

This comment thread is ridiculous, how can any sane person advocate for abolishing the FDA.... (facepalm x10)

Comments for this post are closed