Drug Patents at the Supreme Court

.S. pharmaceutical patent policy is poised for a major review by the U.S. Supreme Court. Later this month, the Court will hear a case, Federal Trade Commission (FTC) v. Actavis, Inc., about a tactic alleged to be illegal by U.S. regulators, socalled "reverse payment" settlements of patent litigation. A maker of a branded drug pays a "generic" drug maker offering a competing, unbranded version of a drug, to abandon its challenge of the branded fi rm's patent. This tactic has also received regulatory scrutiny in Europe ( 2). The Court's ruling promises to reset the innovation/access balance for drugs, whatever the result. We explain the stakes of the case, and how settlements of "secondary" patents affect that balance.

In 1984, the U.S. Congress enacted the Hatch-Waxman Act to balance incentives for drug innovation against robust consumer access. The act enables generics to challenge branded fi rm patents that they believe were erroneously issued (invalid) or impertinent (not infringed), under the theory that such patents ought not deprive consumers of lowpriced generic drugs. Cheaper generic drugs have saved purchasers billions of dollars per year-an estimated $1 trillion in the United over the past decade, according to an industry-sponsored study-which makes such drugs a powerful way to keep down health-care costs ( 3).

Under rules created by the act, a generic must wait until patent expiration or assert that one or more branded patents are invalid or not infringed. That challenge is an act of patent infringement, which often prompts the brand to fi le a lawsuit to prevent the generic's product launch. The outcome of that suitincluding a settlement-determines when the generic fi rm can enter the market.

Since 1984, branded fi rms have substantially stepped up their patenting efforts, in an effort to postpone the loss of exclusivity. In the United States, patents per drug roughly doubled for the cohort of drugs approved between 2000 and 2002 compared with drugs approved between 1985 and 1987 ( 4). Multiple patents with overlapterms result in a longer nominal term of protection. In Europe, the trend is similar ( 1).

Part of this growth refl ects a rise in secondary patents. These provide regulatory protection to ancillary aspects of drug innovation-such as particular drug formulations and compositions-beyond the core, traditional protection, a patent on a novel active ingredient. Secondary patents have become a focus of branded patenting and a major focus for generics' efforts to enter the market before patent expiration. Such patents are viewed as less effective bars to generic entry, compared with active ingredient patents ( 5,6). Secondary patents are thought to be less likely to legal standards of patent validity. Their prevalence, notwithstanding this, refl ects the fact that the Patent and Trademark Office (PTO) generally provides only a limited review of patent applications, which is driven by resource constraints and incentives facing examiners. It is also easier for generic fi rms to avoid infringement by inventing around secondary patents, for example, by devising alternative formulations. For these reasons, generic fi rms' challenges to these patents are likely to be on stronger footing, a point on which we provide empirical evidence below.

Previous work suggests that secondary patents, particularly late-expiring ones, are disproportionately targeted for preexpiration challenge by generic firms ( 7). Such challenges serve a potentially valuable role, subjecting a questionable patent to a second look after issuance. This is important given the relatively cursory review provided by the PTO. Aside from reducing prices sooner, the second look may have a salutary effect on innovation, channeling innovative efforts toward new chemical entities (which typically have "strong" active ingredient patents) and away from incremental improvements (which rely more on secondary patents). That channeling is valuable, provided that litigation tends to validate more deserving innovations.

Reverse Payment Settlements: Disrupting the Balance?

Benefi cial effects of patent challenges can be lost by postchallenge settlements. Consider AndroGel, the drug at issue in the Supreme Court case. AndroGel is a testosterone replacement therapy manufactured by Solvay. In 2003, two drug makers-one of which was later bought by Actavis-sought to introduce generic versions of the drug. Although AndroGel is protected by a patent that expires in 2021, the generics challenged patent as invalid, and not infringed by the proposed generic products ( 8). After the challenges, Solvay sued for patent infringement. In 2006, the parties settled. Under the settlement, generic AndroGel will not be available until 2015, compared with 2006 or 2007 if the patent had been found invalid or not infringed.

The FTC, which brought suit to prohibit the deal, alleges that Solvay made large payments to the generics, in violation of antitrust law, to induce them to abandon patent challenges and to delay entry for 9 years. The FTC argues that such deals, which are increasingly common ( 9), harm drug purchasers by prolonging the period of monopoly profits enjoyed by a branded fi rm. Such settlements have yielded >$12 billion in estimated overcharges since 1993 ( 10).

Branded drug makers argue that a relaxed approach to reverse payment settlements is desirable because higher profi ts promote essential research and development (R&D) that lead to new drugs ( 11). In defense of settlements, representatives of both branded and generic fi rms have pointed to outcomes of lawsuits that do not settle. Often, the branded fi rm wins ( 12,13) 14). The more that settlements pertain to cases the branded fi rm will probably win, the less we ought to worry about reverse payment settlement as disrupting the second level of review provided by patent challenges and litigation ( 15).

These arguments are amenable to empirical analysis ( 16). Previous empirical work ( 10) collected public information about identifi ed settlements with evidence of payment.

For each drug in this data set, collected information about which patents were litigated and settled. Each patent was coded, on the basis of whether at least one claim covered an active ingredient, by using the method outlined in ( 7). If not, the patent was coded secondary ( 17).

For the patents at issue in settled litigation, 89% were secondary patents ( 18). Far from offering generic entry earlier than the presumed likely outcome, a brand win, reverse payment settlements usually limit early entry on patents that would otherwise result in a generic fi rm win. There is therefore serious concern that settlement unduly delays generic entry, harming consumers and potentially distorting R&D incentives.

This conclusion rests on the premise that active ingredient patents are more likely than secondary patents to be found valid and infringed. Some secondary patents are found valid and infringed, and some active ingredient patents are not ( 19). To evaluate this generalization, we collected information on completed patent litigation on all drugs that fi rst became eligible for challenges between 2000 and 2008. The litigation covers 277 patents and 147 drugs. For each patent, we determined the outcome of litigation at the district/trial court level: brand win, generic win, or settlement, and coded the patent as described above.

Of the 48% of cases litigated to completion (not settlement), the branded fi rm nearly always wins a suit asserting an active ingredient patent (92%), but usually loses asserting secondary patents (32% wins) (see the chart). This is consistent with the generalization that active ingredient patents are indeed more likely to be found valid and infringed.

For 52% of cases, settlement is the result. Not all settlements in this data set involve reverse payments; we use these data to assess win rates for different types of patents when litigated to completion. As win rates can be hard to interpret when there is an option to settle ( 20), we examined litigation in an era when settlements were less common, before 2006 ( 21). For this subset (n = 25), all patents were litigated to completion. The branded fi rm won every dispute active ingredient patents and lost nearly all secondary patent litigation (14% wins). These results are consistent with the full sample; generic challengers are likely to prevail on secondary patents if litigated to completion.

We have shown that reverse payment settlements disproportionately focus on secondary patents, and that, consistent with previous qualitative characterizations of these patents, secondary patent challenges are usually won by generics when litigated to completion. This challenges the argument that reverse payment settlements should be tolerated because they are on patents that brands are likely to win. Reverse payment settlements, focused on patents where the generic would have prevailed, interrupt the valuable patent-testing process when generic drug makers challenge patents before expiration. Such settlements are most disruptive in the very context, secondary patents, where this testing is most important.

If the FTC wins its case, we can expect fewer settlements with entry-delaying payments and a consequent increase in early consumer access to lower-priced generic drugs. Reverse payment settlements on secondary patents would be replaced by either procompetitive settlements or litigation, where we would expect a high degree of generic success. Some settlements with payments will survive, because the FTC's proposed rule allows drug makers to justify a payment by proving that it was made for unrelated services, not delay; defensible in light of the avoided litigation expense; or otherwise procompetitive.

If the Court endorses reverse payment settlements, we can expect delayed generic competition until the patent expires, even for drugs whose patent protection is weak. There might seem to be a benefi t, with additional profi ts enjoyed by branded fi rms fueling new innovation. But this argument can be applied to any other conduct, even illegal, such as pricefi xing, that raises branded profi ts.

Our results uncover a further critique, that this bonus accrues mainly to secondary, less important aspects of innovation. Thus, reverse payment settlement (if permitted by the court) can be expected to primarily benefi t follow-on formulations and other forms of innovation covered by secondary patents. We doubt whether distortion in favor of secondary innovation is desirable, particularly given concerns about the slowdown in production of new chemical entities. At a minimum, it is this selective subsidy that settle-ment advocates must explain and defend, not any subsidy to R&D generally.