Manhattan Institute for Policy Research.
Subscribe   Subscribe   MI on Facebook Find us on Twitter Find us on Instagram      


Fix The FDA, Fix Patents, Save Lives

May 13, 2014

By Yevgeniy Feyman

Bringing life-saving drugs to market isn't cheap – the capitalized cost of bringing a typical drug to market is around $1.2 billion and continues to grow every year. And thanks in no small part to heavy-handed FDA regulations, this cost is slated to keep growing, especially for new, highly targeted medicines for diseases ranging from cancer to type 2 diabetes.

Much of the cost of drug development lies in what economists refer to as “fixed” or “sunk” costs – this covers the gamut of everything from machinery to rent to regulatory costs (filing fees and time spent conducting clinical trials for example). This situation creates a classic underinvestment problem – companies invest significant amounts of money to develop a product (with large social benefits) where they will be unable to capture much of the total benefit (social+private); in economic terms, an externality. The way that we overcome the problem of large sunk costs is by granting companies a temporary monopoly on production and sale of a product (in essence, a patent that creates “intellectual property”), that allows society to encourage investment in high-value, high fixed cost products that may not have been developed otherwise. Patents are granted for a 20 year period in all countries that have signed the World Trade Organization's Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs).

The problem is that not all industries are created equal. When Apple unleashes the newest “iWidget,” the engineers who designed it don't need to prove to the FCC or any other government body that the device is “safe and effective.”

But pharmaceuticals aren't smartphones. And ensuring the purity, safety, and efficacy of drugs put out on the market serves a tremendous public interest (particularly in light of the significant information asymmetries between the general public and drug manufacturers). Nevertheless, FDA clinical trial requirements that help prove safety and efficacy in large populations also create a long lag between invention (when patents have to be filed) and commercialization (when the product can be sold). This doesn't happen with other industries – Apple can profit from the iPhone as soon as it is working reasonably well and does well with focus groups. This “commercialization lag” has resulted in an effective patent life of around 12 years in the 90s, which has likely fallen even more since then.

And the shortened effective patent life has potentially dire consequences for U.S. pharmaceuticals. Potentially promising ventures (which may sometimes involve “blue-sky” research), with enormous social value (estimates of the value of a cure for cancer have been as high as $50 trillion) but extensive clinical trial requirements may not get developed or commercialized. Take for example a hypothetical preventative drug for Alzheimer's given to individuals at age 30. Because typically, Alzheimer's develops after age 60, standard clinical trial protocols focused on clinical endpoints would prevent the drug from being developed – the 20 year patent life would be exhausted before any clinical effects could be documented.

In fact, just looking at cancer drugs, researchers from Harvard Law School, the University of Chicago, and MIT found that in one year – 2003 – fixed patent term distortions resulted in a loss of life-years valued at about $89 billion.

So what are we to do?

Fortunately, the research quantifying these distortions also offers a glimpse of the path forward. Generally speaking, the commercialization lag is minimized when drugs are approved based on “surrogate” (non-mortality based) endpoints. (The researchers note that all FDA-approved cancer prevention drugs, which would suffer from under-investment, were approved using surrogate endpoints or through publicly-financed trials.) Surrogate endpoints (this can include molecular changes like the level of the HIV virus in the blood to physical measures like tumor size) tend to shorten the FDA approval process significantly, reducing the cost of clinical trials. To expand the use of surrogate endpoints (especially beyond cancer and HIV), however, requires significant reform at the FDA.

In simple terms, this would mean expanding the agency's “accelerated approval” designation, to most (if not all) drugs (or at least those drugs where there is a well-established link between surrogate markers and clinical results).

But here's the catch – many surrogate endpoints aren't well-vetted. And the value of surrogate endpoints is only relevant insofar as they can accurately predict positive clinical effects.

The other approach, noted by the researchers, would be targeted R&D subsidies. While we currently have a temporary R&D tax credit that helps spur pharmaceutical research, the tax credit has to be renewed each year (creating uncertainty) but it also isn't targeted (that is, it doesn't change the cost of developing a high-value medicine relative to the cost of a lower-value drug). The general idea would be to focus subsidies (perhaps through public financing of clinical trials) on drugs that may not have well-established surrogate endpoints, but may have promise in curing early stage diseases or preventing them altogether.

Both solutions have their share of potential stumbling blocks. For starters, because validating surrogate endpoints can be challenging, increasing federal support for doing so should be a first-order concern. As I've written elsewhere, funding for the National Institutes of Health is critical in this regard, as it is one of the agencies best positioned to help do so.

And while in theory, targeted subsidies can help defray much of the cost of developing high-value drugs, determining which inventions should qualify (generally they should target early-stage diseases and diseases prevention) would likely be on a case-by-case basis (again, increasing uncertainty).

Looking further, while eliminating the commercialization lag would help incentivize firms to invest in areas where the current system discourages spending, the high prices of some drugs would still be an issue. $100,000 cancer drugs, even if developed, can become a substantial cost burden even for some Americans with otherwise generous insurance coverage. (The ACA's annual out of pocket limits will help to some extent, but by mandating coverage of many benefits, the ACA also puts increased upward pressures on deductibles and co-pays, including for high cost medicines, as well as tighter formularies. This is, in fact, the opposite of what you'd like to achieve – routine, one-off costs, should be paid out of pocket, while large costs should be exactly what is covered by insurance.)

One way to address this (that would avoid innovation-killing price controls) is through a tactic initially pioneered by (now) Harvard economist Michael Kremer and endorsed by George Mason University economist, Tyler Cowen – patent buyouts. Under this approach, the government could purchase patents of approved drugs from companies and place them in the public domain, immediately making the drug generic.

Leaving technical details aside (for interested readers, however, the price would be based on second-price sealed-bid auctions), the idea is that the government would pay the inventor the “private value” of the drug, plus some markup. (The sale of the patent would still be entirely voluntary, and the patent holder could always decline the sale.) The private value would be determined through an auction where other companies would be allowed to bid on the patent, and in a small number of random cases, the government would award the patent to the auction winner (this is to ensure that bidders don't bid randomly, but actually reveal their preferences). All in all, for drugs with very large differences between the competitive equilibrium price and the monopoly price (generally drugs with very high social value), this would eliminate the temporary price distortions created by monopoly patents, making life-saving drugs much more widely available (and affordable) than would otherwise be the case.

A few serious caveats: Such a system would necessitate some large government outlays (possibly to the tune of hundreds of billions of dollars, though this cost could be slashed with reforms of the FDA approval process). There is also the problem of follow-on innovations: typically, first in class products will still be on patent when follow on products launch a few years later. With patent buyouts, such drugs would face generic competition immediately, and companies may decide to avoid development new products in this case.

Indeed, the revenues that accrue to first in class drug incentivize additional research. Putting a first-in-class product directly into the public domain after FDA approval, creating an instant generic, would put strong downward pricing pressure on follow-on innovations, reducing incentives for further research. And there's always the danger that public and private insurers would just reimburse for these “generic-first drugs”; the VA, for instance, doesn't cover many newer medicines. This is far from an insurmountable concern, however.

Under Kremer's model, the price paid would likely be greater than the private value of the drug – incorporating a good deal of social value as well, in order to avoid this exact problem. Moreover, for some categories of medicines, where research is especially risky or time consuming, or demand is low (i.e., a malaria vaccine) this approach could effective act as an “advanced purchase commitment”, similar to those used by non-profits today. The long and the short of it is that we are almost certainly underinvesting in new medicines that have enormous social benefits, and driving up their costs once they reach market. As such, thinking carefully about existing patent incentives and FDA regulatory reforms would be necessary to both spur more innovation and assure broad patient access within a reasonable time frame. (as Alex Tabarrok has noted, FDA reform and patent reform could be complementary strategies).

Still, patent buyouts would eliminate much of the monopolistic distortions inherent in all patents, without dampening incentives to innovate.

Of course, this is all easier said than done. Over the long run we need to generate more new drugs, for the people who need them the most. We'll need creative solutions that are sustainable for patients, payers, and industry.

Original Source:



Rolling Stone’s University of Virginia Gang Rape Fiction and the Real World
Heather Mac Donald, 03-25-15

New York Lawmakers' Three Big Blown Chances
E. J. McMahon, 03-24-15

Cruz's Gamble As The Only True Conservative May Pay Off
Diana Furchtgott-Roth, 03-24-15

The Christie Hiatus
Steven Malanga, 03-23-15

Health Independence For Veterans
Avik Roy, 03-23-15

How To Fix Disability Insurance
Scott Winship, 03-23-15

A New Role For Public Broadcasting?
Howard Husock, 03-23-15

Whither Gotham's Transit?
Nicole Gelinas, 03-23-15


The Manhattan Institute, a 501(c)(3), is a think tank whose mission is to develop and disseminate new ideas
that foster greater economic choice and individual responsibility.

Copyright © 2015 Manhattan Institute for Policy Research, Inc. All rights reserved.

52 Vanderbilt Avenue, New York, N.Y. 10017
phone (212) 599-7000 / fax (212) 599-3494