In this third post on reverse settlements, I examine whether traditional legal standards for judging whether a drug patent is valid captures the social value from a drug patent. This is important because the FTC takes the position that reverse settlements extend the expected life of a drug company’s patent as compared to litigation over patent validity under traditional legal standards for judging patent validity. I conclude that those standards may seriously undervalue the social benefit of drug patents, even invalid ones. First, patent validity standards do not appreciate that all drug patents – valid or not – are necessary to compensate drug companies for conducting clinical trials, which are half the cost of all R&D and are socially valuable even if the drug patent is not valid. Second, profits from a drug patent – even if the patent is invalid – sustains research on a large number of other drug patents – which may be valid. That cross-subsidization suggests a branded drug company ought be judged on their portfolio of patents, not on individual patents. Traditional patent validity standards fail to do that.
To be clear, I understand that patent law standards for validity are not perfect, and I do not expect them to be. What I suggest is that there is a worse fit between patent validity standards in the pharmaceutical industry than in other industries. This is relevant for antitrust analysis because, if patent validity standards undervalue drug patents, then eliminating reverse settlements undervalues them even further since these settlements putatively extend the expected patent life of drug patents.
Let me begin with my first substantive claim: all drug patents, whether valid or not, are socially useful. All drug patents encourage production of useful public goods, specifically information about whether a drug works. The main reason is that, unlike patents for other products, a patent for a drug is obtained before research on the drug is completed. Specifically, drug companies obtain patents on molecules after lab and animal testing, but before clinical test, i.e., testing on humans.[1] Yet human testing is roughly half the cost of all drug R&D (DiMasi, Hansen, & Grabowski, 2003). Moreover, the results from clinical trials, which are made public as part of the drug approval process, are a public good. Once a drug is shown to be effective, everyone knows whether the molecule is medically valuable. If a company did not have a patent that could prevent other companies from producing that molecule, it would never conduct the trial in the first place!
The other thing that distinguishes drug R&D, specifically clinical trials, from R&D for other products is that clinical trials are socially valuable even if the patents covering a drug is invalid. It is socially valuable to learn whether a drug is safe and effective and that value is positive whether the relevant patent is valid or not. Indeed, Ben Roin at Harvard Law School notes that a large number of molecules in the public domain – either because they were not patentable or their patents expired – may be valuable but are untested and unexploited because no company can recoup the costs of doing research on them.[2] Further evidence comes from the Orphan Drug Act, which rewards drug companies that demonstrate via trials that a molecule is safe and effective at treating an orphan disease with up to 7 years of market exclusivity, even if the molecule is already in the public domain.[3] The Orphan Drug Act is thought to have successfully increased the number of treatments for orphan diseases.[4] If market exclusivity for drugs that are in the public domain have social value, so is market exclusivity for drugs that may have invalid patents.
It is true that knowing a drug works may be more valuable than knowing a drug does not work, and some drugs with invalid patents may not be safe and effective treatments. But the manner in which generic drug companies select patents for Paragraph IV challenges – the Hatch Waxman provision that gives rise to patent suits resolved via reverse settlements – suggests that reverse settlements largely cover drugs that work, thus drugs for which trials are particularly valuable. All Paragraph IV challenges are filed against drug approved by the FDA, i.e., found to be safe and effective. Moreover, Scott Hemphill and Bhavan Sampat demonstrated in a 2011 study that sales are an important predictor of Paragraph IV challenges.[5] And sales tend to follow drugs that work well rather than drugs that do not. This should not be surprising: generics are also in it for profit. Drugs that give the greatest value from Hatch-Waxman’s 180 market exclusivity period for the first-filing generic are exactly those drugs that are most profitable to the prioneer drug company.
One criticism of this line of argument is that patents may reward drugs companies “too much” for trials, i.e., give them more money than is required to get them to do trials on the drug in question. For drugs that are subject to Paragraph IV challenges, that is almost surely true: the profits they generate are much larger than the $250 million or so it takes to conduct 3 phases of trials. Of course, there are many drugs for which trials are conducted and the drug is not approved. Cost estimates that account for the fact that profits from drugs that are approved pay for trials for drugs not approved range from ~$800 million – $2 billion per approved drug [6]. Likewise many drugs for which trials are conducted and drugs are approved, but drug does not make enough to cover costs of development. Cost estimates here are unknown. If we assume the median drug does not cover costs of development [7], then costs could be double, or as high as $1.6 billion – $4 billion. Now the profits on drugs subject to Paragraph IV challenges seem less excessive. Frankly, I don’t know how much the trial cross-subsidy must be. And I am confident that total profits in expectation are larger than that – because they also have to cover the cost of pre-trial, pre-patent research. Plus they have to cover other costs, including, controversially, marketing – a topic I’ll address in a moment. But I don’t know if the excess profits from patents go so far beyond that that drug companies may be supra-competitive profits in the market for R&D. The Office of Technology Assessment (1994) estimated that the return on assets for drug companies are only 2-3% higher than their 10% real cost of capital.
Another concern with drug patents is that, valid or not, they support marketing that inflates demand. But from a welfare perspective, there are pros and cons. On the one hand, it may get individuals for whom utilization is not appropriate to consume. One the other hand, it offsets high monopoly price. The question is whether marketing causes people to consume the wrong drug because of quality or whether marketing causes individuals to purchase a drug that was inappropriate because of price. If the latter, marketing could be welfare enhancing.
A third concern with supporting even invalid patents to encourage clinical testing is that drug company funding of research may not truly produce a public good. Specifically, prior research has found that drug company funding of research causes research outcomes to be “biased” in favor of positive treatment effects (Bero, Galbraith, & Rennie, 1992; Cho & Bero, 1996; Djulbegovic et al., 2000); (Barnes & Bero, 1998; Bekelman, Li, & Gross, 2003; Bennett et al., 2010; Davidson, 1986; Djulbegovic et al., 2000; Friedberg, Saffran, Stinson, Nelson, & Bennett, 1999; Friedman & Richter, 2004; Kjaergard & Als-Nielsen, 2002; Rochon et al., 1994; Turner & Spilich, 1997; Yaphe, Edman, Knishkowy, & Herman, 2001). But we don’t know the baseline: it could be that the non-funded research is wrong. It is our priors that lack of funding means no bias that yields the inference of bias due to drug company funding. It is also possible that lack of funding means lower quality so the non-funded research is erroneous. Moreover, we don’t know the consequence of that bias. For example, it could be that misinformation about quality is what is required to overcome monopoly price.
Finally, the quality of evidence on bias due to drug company funding is not tremendous. There are quality differences between drug company funded studies and other studies, not always disfavoring drug company funded trials (Djulbegovic et al., 2000). That opens the possibility that there are quality differences not captured by observable variables. So the finding of bias could be due to selection, specifically unobserved quality differences favoring drug companies. One source of such selection could be attributable to differences in the researchers that drug companies and non-drug companies (often the government) select to conduct studies. Just as drug companies may select scholars who have priors or techniques favoring drug company products, the non-drug companies could be doing the opposite. It is not obvious which type of trial is more biased relative to the truth.
My second claim is that, for antitrust purposes, drug companies should be judged on the validity of patents across all drugs in their portfolio, not the validity of individual patents. Because even an invalid drug patent may subsidize socially valuable research on another one, the value of that patent is not fully captured by its validity alone.
Drug companies typically conduct trials on a number of drugs and even sell multiple approved drugs on the market. Carrying a number of drugs reduces the variability of a drug company’s profits. This smoothing of profits reduces the cost of capital to a drug company. Investors demand a higher rate of return on more risky investments. Offering investors a portfolio of drugs offers lower risk investment than offering just one drug; investors this demand a lower return from a company with a portfolio of drugs than a company with an individual drug ceteris paribus. Thus holding portfolios of drugs enable drug companies to raise capital at a lower cost.
Carrying a portfolio of drugs also allows drug companies to lower capital costs by partly self-finance investment in innovation. Raising capital for R&D is difficult because companies typically have more information about the prospects of their R&D projects than lenders or investors. This asymmetry makes lenders and investors reticent to offer capital except at high rates of return. An alternative to paying those high rates is for the company to finance R&D through retained earnings. Holding a portfolio increases the odds that at least one drug generates positive profits that can be retained for financing purposes.
If we combine the the fact that drug companies carry portfolios with the fact that, adjusting for risk, drug companies do not make substantially supracompetitive profits (OTA 1994), we can infer that drug patents that generate high profits cross subsidize R&D on unprofitable patents on approved drugs and also patents on unapproved drugs. Paragraph IV challenges threaten this financing system because they target drugs with high sales. That such challenges might focus on relatively weak patents (Hemphill & Sampat 2011) does not affect my conclusion because patent invalidity does not affect the value of profits from that patent or the ability to use those profits to fund research on other patented drugs.
In short, an invalid patent may finance socially valuable research on a valid patent. Thus the first patent’s social value depends not only on its own value, only partly captured by its validity, but also the value of the patent it cross-subsidizes, again only partly captured by its validity.
A natural criticism is that allowing in valid patents into a drug company’s portfolio increases the overall level of invalid patents in that portfolio. That may be true. But the position taken by the FTC and some legal scholars is that reverse settlement should be per se invalid. I.e., even one invalid drug patent should be quashed. That rules out the possibility that a portfolio of drug may have some – but not all – invalid patents, that the portfolio would not exist (or have as many valid patents) without the invalid patents, and that the portfolio as a whole is socially valuable. This possibility suggests a role for rule of reason analysis, and one that is not confined to just the patent subject to a Paragraph IV challenge. It certainly counsels against a per se rule that reverse settlements are invalid.
[1] If companies waited longer, the publication of lab and animal testing results might make the molecule unpatentable as after clinical testing.
[2] Roin BN (2008) Unpatentable drugs and the standards of patentability. Tex. L. Rev. 87:503.
[3] An orphan disease is one that afflicts fewer that 200,000 individual per year.
[4] Fill in cite.
[5] Hemphill CS & Sampat BN (2011) When Do Generics Challenge Drug Patents? Journal of Empirical Legal Studies 8(4):613-649.
[6] DiMasi et al. (2003) give the ~$800m estimate. See also Adams, C. P., & Brantner, V. V. (2010). Spending on new drug development1. Health economics, 19(2), 130-141. doi: 10.1002/hec.1454 (confirming DiMasi et al. estimate). Adams and Brantner suggest that for some diseases the costs may rise to $2b. Adams, C. P., & Brantner, V. V. (2006). Estimating The Cost Of New Drug Development: Is It Really $802 Million? Health Affairs, 25(2), 420-428. doi: 10.1377/hlthaff.25.2.420. A McKinsey Consulting estimate suggests the costs could be as high as $2 billion per drug. [Insert citation.] But see Light & Warburton (2011) (estimating cost is just $43.4m per drug).
[7] See Grabowski, H. G., & Vernon, J. M. (1994). Returns to R& D on new drug introductions in the 1980s. Journal of Health Economics, 13(4), 383-406. doi: https://dx.doi.org/10.1016/0167-6296(94)90010-8 (esp. Fig. 5).
References cited in text
Barnes, D. E., & Bero, L. A. (1998). Why Review Articles on the Health Effects of Passive Smoking Reach Different Conclusions. JAMA: The Journal of the American Medical Association, 279(19), 1566-1570. doi: 10.1001/jama.279.19.1566
Bekelman, J. E., Li, Y., & Gross, C. P. (2003). Scope and Impact of Financial Conflicts of Interest in Biomedical Research. JAMA: The Journal of the American Medical Association, 289(4), 454-465. doi: 10.1001/jama.289.4.454
Bennett, C. L., Lai, S. Y., Henke, M., Barnato, S. E., Armitage, J. O., & Sartor, O. (2010). Association Between Pharmaceutical Support and Basic Science Research on Erythropoiesis-Stimulating Agents. Arch Intern Med, 170(16), 1490-1498. doi: 10.1001/archinternmed.2010.309
Bero, L. A., Galbraith, A., & Rennie, D. (1992). The Publication of Sponsored Symposiums in Medical Journals. New England Journal of Medicine, 327(16), 1135-1140. doi: doi:10.1056/NEJM199210153271606
Cho, M. K., & Bero, L. A. (1996). The Quality of Drug Studies Published in Symposium Proceedings. Annals of Internal Medicine, 124(5), 485-489.
Davidson, R. (1986). Source of funding and outcome of clinical trials. Journal of General Internal Medicine, 1(3), 155-158. doi: 10.1007/bf02602327
DiMasi, J. A., Hansen, R. W., & Grabowski, H. G. (2003). The price of innovation: new estimates of drug development costs. Journal of Health Economics, 22(2), 151-185.
Djulbegovic, B., Lacevic, M., Cantor, A., Fields, K. K., Bennett, C. L., Adams, J. R., . . . Lyman, G. H. (2000). The uncertainty principle and industry-sponsored research. The Lancet, 356(9230), 635-638.
Friedberg, M., Saffran, B., Stinson, T. J., Nelson, W., & Bennett, C. L. (1999). Evaluation of Conflict of Interest in Economic Analyses of New Drugs Used in Oncology. JAMA: The Journal of the American Medical Association, 282(15), 1453-1457. doi: 10.1001/jama.282.15.1453
Friedman, L. S., & Richter, E. D. (2004). Relationship Between Conflicts of Interest and Research Results. Journal of General Internal Medicine, 19(1), 51-56. doi: 10.1111/j.1525-1497.2004.30617.x
Kjaergard, L., & Als-Nielsen, B. (2002). Association between competing interests and authors’ conclusions: epidemiological study of randomised clinical trials published in the BMJ. BMJ, 325(7358), 249. doi: 10.1136/bmj.325.7358.249
Rochon, P. A., Gurwitz, J. H., Simms, R. W., Fortin, P. R., Felson, D. T., Minaker, K. L., & Chalmers, T. C. (1994). A Study of Manufacturer-Supported Trials of Nonsteroidal Anti-inflammatory Drugs in the Treatment of Arthritis. Arch Intern Med, 154(2), 157-163. doi: 10.1001/archinte.1994.00420020059007
Turner, C., & Spilich, G. J. (1997). Research into smoking or nicotine and human cognitive performance: does the source of funding make a difference? Addiction, 92(11), 1423-1426. doi: 10.1111/j.1360-0443.1997.tb02863.x
Yaphe, J., Edman, R., Knishkowy, B., & Herman, J. (2001). The association between funding by commercial interests and study outcome in randomized controlled drug trials. Family Practice, 18(6), 565-568. doi: 10.1093/fampra/18.6.565