Skip to main content

Infectious Diseases

Fun With Priority Review Vouchers

So in that last post, I mentioned a Priority Review Voucher as a likely reason for Marin Shkreli’s latest machinations, and that makes me think that I should explain what the heck a Priority Review Voucher is, since it’s an obscure topic if you’re not into the ins and outs of regulatory drug approval.

It’s a relatively recent invention, going back to the FDA Amendments Act of 2007. The idea, proposed in this 2006 paper, was to provide an incentive for drug development in areas whose markets would not normally invite enough investment. The first category was neglected tropical disease – the agency provided a list of diseases, and the provision was that if a company had a new therapy approved against one of these, that they would also receive a voucher good for “priority review” of a future drug application. The agency bound itself to making that review happen within six months, a potential savings of several months of regulatory hold-up.

Most importantly, these vouchers were fungible: once issued, they could be resold to the highest bidder. Now, that may seem a bit odd, because you’d think that a company that had been awarded one of these would want to hold on to it. But it depends on their cash flow –  a small company may well need the money that the sale of the voucher could generate more than they need to hold onto it in the hopes that they’ll be able to get another drug in for review any time soon. Meanwhile, a larger company, especially in a competitive market, may well find that paying quite a bit for a priority review is still worthwhile.

Consider, also, that the idea is for the vouchers to be incentives for R&D. If they weren’t transferrable, they’d still be worth something, but nowhere near as much as if there were a secondary market. And several of them would surely expire unused, if the original companies that obtained them weren’t able to use them, sell them, or trade for them. So when the FDA issues one, its committing itself to a priority review of something in the future, the identity to be determined by who wants a priority review the most.

That’s exactly how it’s worked out. The FDA had sixteen tropical diseases on its list (things like blinding trachoma, leishmaniasis, dengue, guinea worm disease), and Congress has added some viral diseases (like Marburg and Ebola). More recently, the FDA also added Chagas and neurocysticercosis to the list. In 2012, a very similar program was put in for rare pediatric diseases (those primarily affecting people 18 and younger, and present in 200,000 patients or fewer). For a couple of years, there were significant differences between the two types of voucher (for example, the tropical disease ones required a year’s notice to the FDA and could be resold only once), but these differences were eliminated in 2014. Both types now can be redeemed with 90 days of notice, with unlimited resale. The FDA does charge a fee for priority review, several million dollars extra, but that’s not going to slow anyone down.

It doesn’t slow anyone down, because, as it turns out, these vouchers are worth hundreds of millions of dollars each. That link has a chart of the resale values, which have done nothing but climb. A recent high-profile use of one was Sanofi-Aventis/Regeneron beating Amgen to market in the US in the PCSK9, although Amgen had made it first in Europe. It was only a month’s head start, but in a market that large and competitive, it may well have been worth it (and sales figures are being watched closely by many others to see if it looks like it was).

These programs are an unusually clear and direct example of a pure economic solution being applied by a large government agency. Making the vouchers completely resellable is the key: that creates a market which then asks the basic question of capitalism: “What’s it worth to you?” The only way to find out how much money some good or service is actually worth is to see what people will pay for it, and so far it looks as if drug companies will pay quite a bit to have their drugs reviewed quickly. The incentives to develop drugs for otherwise-neglected diseases, then, are much stronger, and it’s not the government that has to offer the cash – drug companies themselves are paying for it. Overall, I think it’s an excellent idea, one that uses market mechanisms to correct a market failure.

Of course, there are complications. The biggest question is which diseases get this sort of special treatment. Someone has to make that call, and reach in and put in the bypass wiring from outside – there’s no market mechanism for that, because if there were, we wouldn’t be having to mess around with things like priority review vouchers in the first place. There are opposing mistakes to be made. You could designate a disease that, honestly, isn’t enough of a problem to make the award of one of these vouchers appropriate. (That’s the mistake that I complain about when the FDA awards market exclusivity for taking an older drug through the regulatory process – I think the payoff is disproportionate and distorting). On the other end of the scale, a disease like malaria has a worldwide toll in human suffering, death, and economic hardship that is worth more to alleviate than a single priority review voucher can cover. So not only does someone have to make the call of which diseases get preference, but there could be another call to make about the degree of preference. Should a powerful new malaria approach get two or three priority reviews awarded, for example?

But overall, I think that the idea is sound. Now, whether Martin Shkreli and KaloBios should be handed a PRV for a Chagas drug that’s already being used to treat Chagas disease. . .now that’s another question.

32 comments on “Fun With Priority Review Vouchers”

  1. Andrew says:

    “And several of them would surely expire unused, if the original companies that obtained them weren’t able to use them, sell them, or trade for them.” – not that it changes the overall point, but the vouchers don’t expire.

    1. Bosh says:

      True, but I read Derek’s comment to mean that if the vouchers were not fungible – that is, if they were inextricably tied to the company awarded them – then they would effectively expire should that company collapse, as biotechs have been known to do.

  2. PorkPieHat says:

    Surely the FDA can choose NOT to award a priority review voucher for a drug that’s ALREADY being used to treat the disease, no? I dont get it, in the Shkreli case. Must say, he knows how to push the envelop of what is possible in this industry, and in doing so, exposes the larger problems. As such, he may end up being good for us all, if something is to be done about those very problems (pricing loopholes, etc).

  3. Chrispy says:

    Maybe it is time to admit that our drug approval system is fundamentally flawed. If a six-month review is possible, then perhaps a six month review should be standard, voucher or not. Every month that the FDA delays a drug approval through its own Byzantine process is a month of patients getting denied therapy. Our process is too expensive, too slow, and too opaque

    1. Sean says:

      I’m not sure that makes sense. Just because the FDA has the resources to process a few applications per year faster, doesn’t mean they have the resources to process every application faster.

      1. Erebus says:

        But surely you must see the need for a more standardized, more predictable, and more transparent process?

        To go off on a bit of a tangent: The way I see it, the aim should be to expedite review inasmuch as possible, and to value well-defined safety parameters over vague and amorphous “efficacy” parameters. Frankly, I’m not convinced that the market can’t sort out the issue of efficacy on its own. (With help from the scientific publishing organizations, the NIH as a body which often makes recommendations for standard treatments, etc.)

        1. LZ says:

          Frankly, I’m not convinced that the market can’t sort out the issue of efficacy on its own.

          Given the popularity of useless nutritional suppliments, I’m pretty sure it can’t.

          1. Emjeff says:

            That’s a somewhat spurious argument, since nutritional supplements are bought and used by lay-people, and prescription drugs are “bought” by insurance companies (who are free to not cover something if the data are not to their liking). Moreover, FDA does not really have the expertise in many therapeutic areas to be making the calls in this space. What they can do a reasonable job of is evaluating safety.

        2. Ex-academic says:

          How do you evaluate safety without taking efficacy into account?

        3. tangent says:

          @Erebus ‘and to value well-defined safety parameters over vague and amorphous “efficacy” parameters. Frankly, I’m not convinced that the market can’t sort out the issue of efficacy on its own’

          Innocent question: what is just so vague and amorphous about efficacy? Are relevant endpoints that hard to come by? The market-driven FDA-free efficacy trials are going to need some endpoints still.

          Or are you envisioning that we don’t do RCTs for efficacy, we just put drugs on the market with their safety risks documented? Because no. Doctors, love ’em, but they’re bad empiricists like this. The best we’d have, heaven help us, is the actuaries running retrospective analyses over medical records. Relative to the current system, that will degrade our standard of care and harm patients.

          1. Erebus says:

            Extensive “retrospective analyses over medical records and patient outcomes” is exactly what I’d propose — and it should go without saying that such things are exactly what insurance agencies and medical professionals would demand, under the circumstances.

            …But I don’t think that this is as undesirable as you’re making it out to be. Post-market data analysis can assess benefits and harms across an extremely large population, can be performed very cheaply and quickly, and can even uncover relevant effects (e.g. drug interactions) that the FDA might miss.

            The FDA was founded as a consumer protection agency, and it can be quite effective, if not entirely effective, at ensuring drug safety. It simply cannot be effective at determining whether or not a drug candidate is efficacious. In trying to do police efficacy, it has let drug development costs — and, importantly, timeframes to regulatory approval — creep out of control. This is very much to the detriment of patients, the pharmaceutical industry, and even the public at large.

            Drug safety testing is amenable to standardization. The agency could use something along the lines of a point system, which would apply to every drug candidate without exception. (Side effects graded by likelihood and severity, and then assigned a certain number of points, with “point allowances” for different indications — for instance, 150 for the treatment of mild anxiety, 2000 for aggressive cancer treatments and “last line of defense” antibiotics.)
            …This sort of thing can be easy, transparent, reproducible, and fair. Such a thing would be be a boon to drug development.

            Testing efficacy is precisely the opposite — it is impossible to standardize, it is never straightforward, and the rules vary wildly in each individual case, which makes the FDA look inept and mercurial. Certain indications are de facto prioritized over others — for instance, efficacy trials for acute disease treatments cost vastly less than the long and expensive trials demanded of Alzheimer’s treatments — and nobody has any idea at all how to run meaningful efficacy trials for anti-aging treatments.
            …In general, “efficacy” trials keep getting longer and more expensive over time. This is the single greatest burden facing drug development today. If we want a faster, more innovative, and more patient-oriented system, we’ll want to have the FDA’s regulators focus on safety, which is the primary mission of their organization, and leave efficacy to the industry, to academia, and to insurance groups and other payers.

      2. Anonymous says:

        And perhaps that is something to be solved by the market as well. If the prioritized review can be enabled by paying an additional fee, how much to prioritize regulatory reviews without the voucher?

      3. John Thacker says:

        Sure, but it implies that there may be a large benefit to giving the FDA more resources so that they can process all drugs faster. If the benefits of processing drugs faster are on the level of “hundreds of millions of dollars” per drug, and the costs of hiring more people at the FDA (or whatever is necessary to improve the process) is lower, then there can be a societal benefit.

        The FDA should be shouting this number to the rooftops and using it to argue for a larger budget.

        1. Mark Thorson says:

          FDA has a budget of about $5B of which about $2B comes from user fees from pharmaceutical companies. By comparison, NIH has a budget of about $30B. Sure, researching new science for drugs is nifty, but maybe a billion or two could be spared from research to grease the wheels for getting stuff that we’ve already discovered to market. In a sense, that’s part of research too.

    2. M says:

      You could certainly make a case that all approvals should take the same time, but it’s not self-evidently true.

      If the FDA gets two applications at the same time, one for a cancer treatment where no therapy exists and one for an analgesic that targets a well served population, I’d expect them to approve the cancer treatment much more quickly. It’s both a lower bar and more important.

  4. Jacob says:

    STAT had a nice review article covering Priority Review Vouchers and issues with the current system.
    Most prominent is the issue Derek mentioned in closing — PRVs are being given out for *existing* drugs, *already used for therapy* simply because they are being put through FDA approval for the first time. There are more of these than I would have guessed, but it makes sense — many of these diseases don’t occur in the US, so why bother with the FDA?
    I agree that the incentive system is a nice fix for a market failure, but we have to make sure we reward *new therapeutics* rather than simply *new approvals*.

  5. Jacob says:

    STAT had a nice review of this problem recently. Posting a link gets my comment deleted, so just google the title “How a system meant to develop drugs for rare diseases broke down.”

    In it, they highlight the issue Derek mentions in closing as the primary problem with our current system. Priority vouchers need to reward new therapeutics, rather than simply new approvals of existing therapies.

  6. Morten G says:

    Okay, either I’m an idiot or this is a moronic system. Ahhhh! Now I get it. I’m an idiot.
    One therapy approval wins one express review which can be used in the approval process of a blockbuster drug. And it should obviously be sellable since a company specialised in tropical diseases is unlikely to derive much value from an express review. Gotcha.
    Aren’t there a limited number of therapies in the world that you could perform the KaloBios trick with? I think miltefosine for Leishmaniasis is only approved in India. That might be one (miltefosine is so toxic though). Others? If there’s only a few then it’s a very limited ressource and the issue will resolve itself.

    1. Mark Thorson says:

      Sapropterin is approved in Japan for autism, but not in the U.S. I’m skeptical that it really works, but you could probably sell a lot of it to desperate parents. Insurance companies might balk, though. (I think they should.) It is approved in the U.S. for treating certain forms of atypical phenylketonuria.

      On the other hand, it’s completely harmless. Pretty much all of the adverse side effects are due to the acid content (it’s the HCl salt of tetrahydrobiopterin).

      There’s a German study from about 30 years ago which seems to show greater efficacy for sepiapterin (tetrahydrobiopterin precursor) in the treatment of atypical PKU, probably due to its much better ability to cross the blood-brain barrier. I don’t know why sepiapterin has not been commercialized for the treatment of atypical PKU, other than sapropterin got a head start. Tetrahydrobiopterin deficiency is the cause of most forms of atypical PKU, so most of the research studies have used sapropterin.

      Sepiapterin is greasier and much more stable than sapropterin. You might be able to get away with a less aggressive acid salt, like maybe a tartarate instead of HCl. I think those kids with atypical PKU could benefit from replacing sapropterin with sepiapterin, but I don’t know how to make that happen.

      1. Morten G says:

        The neglected childhood diseases PRV is more restrictive in some senses so I don’t think that would fly. Maybe. From Wikip.:

        “The act extends the voucher program to rare pediatric diseases, but only on a trial basis. One year after the third pediatric voucher is awarded, no other pediatric vouchers may be awarded. The extension to rare pediatric diseases was championed by Nancy Goodman of Kids v Cancer.

        The pediatric voucher program includes several changes that had been sought for the neglected-disease voucher program, but apply only to pediatric vouchers. First, the pediatric treatment developer can ask the FDA in advance for an indication of whether the disease qualifies as a rare, pediatric disease. Second, the pediatric voucher can be transferred an unlimited number of times, whereas the neglected-disease voucher can only be transferred once. Third, the pediatric voucher user needs to notify FDA 90 days prior to using the voucher, rather than 1 year for the neglected-disease voucher. Fourth, the pediatric voucher winner risks having the voucher revoked if the treatment is not marketed within a year. Fifth, the pediatric voucher winner must report to FDA about use of the pediatric treatment within five years of approval.”

        It might be good to run both compounds through clinical trials, considering your scepticism about efficacy. But since you need FDA stamp of approval for your run at getting a PRV they could deny it based on it not being a novel therapy.

        I think the fix would be easy. Make any treatments approved anywhere in the world before 2007 ineligible the PRV program (for same indication obviously). Any “novel” treatment containing an API approved before 2007 should be evaluated by the FDA before submission for eligiblity in the same way as the rare pediatric diseases (for combination treatments etc). Any companies trying to use the loophole should be able to submit before it gets signed into law so they can’t cry foul. Of course that would mean that KaloBios gets it’s Chagas voucher but at least that brings more attention to the programme.
        PS AbbVie overpaid at $350 million. The Duke researchers calculated that the accelerated FDA review could be worth more than $300 million but this was for a $3 billion drug. You can’t know a priori if you get approval or what the value of the drug will be. So only retrospectically could the PRV be worth more than $300 miilion (defining worth as drug sales not what some bozo pays for the PRV).

  7. Derek Freyberg says:

    OK, so a number of us think Shkreli is an awful person (I do), and that the PRV system is warped (maybe).
    But how come none of the big pharmas, full of MBA types, thought of picking up KaloBios to get a PRV? It would be a LOT cheaper than buying it from good ol’ Martin.

    1. M says:

      The reputation of a pharmaceutical company, at least with doctors, advocacy groups, and the FDA, matters to them as professionals but also in a business sense. Novartis has a lot to lose if the FDA decides they aren’t trustworthy–a few months delay on a future approval will cost a lot more than you could ever get gouging these niche markets, even including the voucher value.

      Shkreli clearly doesn’t care about this and is less exposed in a business sense, so he’s immune to the norms of the industry. Thus an incentive system that seemed to make sense before now looks ridiculous.

      1. AVS-600 says:

        There’s nothing that says you HAVE to price-gouge if you have market exclusivity. If whatever famous/name-brand pharma company wanted the voucher but was worried about that publicity, they could immediately transfer the rights back to whoever was distributing it before (or even distribute the drug at-cost/for free; it sounds like it wouldn’t be that expensive compared to the value of the voucher).

  8. Unsure says:

    I’m a bit confused how a PRV could be used to speed review of the S-A/Regeneron drug targeting PCSK9. How does a drug treatment for a high LDL condition warrant priority review? This seems incongruous with the FDA’s stated purpose of using PRVs toward expedited reviews of rare/neglected diseases.

    Sorry if this is obvious and I’m missing something.

  9. Me says:

    So we should be approving drugs based on safety, then measuring efficacy with post-approval studies?

    That’s basically alternative therapies – and lack of efficacy is a ‘pharma conspiracy’…..

    There is a certain question of ethics…..

    1. Erebus says:

      That’s the way it was done until the Thalidomide scandal. That’s how it should be done. The FDA is simply not well-equipped to judge efficacy; in trying to do so, it’s stifling valuable innovation, delaying access to treatments, trapping rare diseases and unusual indications in regulatory no-man’s-land, and, ultimately, harming many millions of patients.

      As somebody mentioned earlier, the NIH has a far larger budget than the FDA. They are very well equipped to review data/outcomes and issue recommendations to insurance companies and medical professionals. Academia should also play a role — as can curious physicians, actuaries, and the scientific press. Whatever the downsides of this sort of arrangement, I am convinced that the status quo is far worse.

      1. Me says:

        There’s a reason why the Thalidomide scandal changed things….. and that was a safety issue anyway, so I don’t see your point.

        In Europe, the various health authorities have systems that perform these reviews – in the US the system is a little different, being a little more market-based. This is why a lot of expensive drugs are not reimbursed on government-run healthcare schemes in Europe – but they are approved and can be purchased privately outside of these schemes. The drug companies then buy a lot of space in the press where they run stories about how access to drugs is being rationed etc. You are basically asking for a US-version of NICE to run Healthcare Technology Assessments.

        ie it’s already done

        1. Erebus says:

          Let me break it down for you:
          -The reaction to Thalidomide was unreasonable. It had nothing to do with the FDA’s core mission. It was a mistake.

          -Notably, Thalidomide was not approved in the USA for safety reasons. Efficacy had nothing to do with it. (One could note that Thalidomide was highly efficacious, if a bit of a dirty drug.) The regulations which were in place at that time worked adequately.

          -The FDA is not well equipped to regulate drug efficacy. Its track record is dismal — particularly with respect to trial costs & timeframes to approval, which are frankly out of control & getting worse. What’s more, it’s mercurial — some would say openly corrupt. Surely what’s not working should be fixed, and old mistakes should be corrected?

          -In my opinion, the FDA should get “back to its roots” as a consumer safety agency, and leave efficacy concerns to organizations that are broader in scope. Besides, there is less harm than you think in releasing safe, less-than-optimally-effective drugs, particularly in this modern Internet age. Treatments are analyzed all the time. The NIH and other respected bodies make treatment recommendations that physicians follow.

          …All that aside, I don’t quite understand what you’re trying to say about Europe. And your claim that the USA’s current regulatory process is “more market based” is a real head-scratcher.

          1. Me says:

            Regulators generally don’t require a strong efficacy endpoint – the ‘corruptness’ of the FDA is generally considered that it allows drugs that aren’t particularly efficacious on the market. But in my view it’s no different to any other agency – just in semantics. Generally, the hurdle for agency approval is much lower than the hurdle for high level reimbursement. There are plenty of post-approval registries and studies being run all the time to assess efficacy – but in the end, efficacy is relative and endpoints need to be reached.
            It seems your opinion is that we just approve anything as long as we say it’s safe, and then we wait 20 years for enough post-approval data to come along. Then the patients that got the good stuff and have good outcomes are fine, and the ones that got the duff ones go around and sue all the medics that prescribed them the inefficacious treatments. And then efficacious treatments are ignored in place of ‘new’ treatments that are unproven…..

            I have a jar of snake oil for you

        2. Erebus says:

          “Regulators generally don’t require a strong efficacy endpoint”. Okay, so why the onerous clinical trial demands? What need, then, for the crushing expenses they incur? Surely these things are, to say the very least, inefficient? Is it not likely that they are actively hindering drug development? (The question to ask is: At what cost are we filtering out drugs that don’t meet their efficacy endpoints?)

          “Efficacy is relative” — I agree, and, as I’ve stated earlier, this leads to the de facto regulatory prioritization of certain indications over others. This is, emphatically, not a good thing.

          Your “wait 20 years” remark is basically a strawman argument. It wouldn’t take nearly so long. Besides, we wait as long as 15 years for drug approval now — primarily so that drugs meet their “efficacy” endpoints — and yet the FDA nevertheless “allows drugs that aren’t particularly efficacious on the market.” So what then? Why must “endpoints need to be reached” prior to drug approval? Why not allow broader access to potentially meaningful treatments, particularly when they are proven safe?
          …Besides, did you read the article? It seems the FDA’s efficacy requirements are inherently meaningless, as they buckle under lobbying and political pressure.

          Simply put: The FDA is doing a catastrophically poor job of policing drug efficacy. Alternatives should be considered.

          As for “Snake oil”: It’s funny that you have such a touching faith in the FDA, and yet lack all faith in physicians, in their professional organizations, and in the relevant legal, medical, and academic establishments.

          1. Me says:

            Think it would be beneficial if you described the pathway a little… how do you approve a new treatment when others exist that have a decent record?

            ‘Proven A’ + ‘New B’? Then revert to B if you see A+B is beneficial to A only.

            Or withhold A and give B and hope the patient doesn’t die?

Comments are closed.